The Effect of Conducting Gesture On Expressive-Interpretive Perfo

University of South Florida
Scholar Commons
Graduate Theses and Dissertations Graduate School
2005
The Effect of Conducting Gesture on Expressive-Interpretive

Performance of College Music Majors
Ronald Wayne Gallops
Follow this and additional works at: https://scholarcommons.usf.edu/etd
Part of the American Studies Commons
Scholar Commons Citation

Gallops, Ronald Wayne, "The Effect of Conducting Gesture on Expressive-Interpretive Performance of
College Music Majors" (2005). Graduate Theses and Dissertations.
https://scholarcommons.usf.edu/etd/2885
This Dissertation is brought to you for free and open access by the Graduate School at Scholar Commons. It has
been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Scholar
Commons. For more information, please contact scholarcommons@usf.edu.
The Effect of Conducting Gesture
on Expressive-Interpretive
by
A dissertation submitted in partial fulfillment

of the requirements for the degree of
Doctor of Philosophy
School of Music
College of Visual and Performing Arts
Major Professor: Jack Heller, Ph.D.

David Williams, Ph.D.
Michael Robinson, D.M.A.
Sheila Woodward, Ph.D.
Date of Approval:
June 16, 2005
Keywords: Music, Conducting, Gesture, Interpretation, Perception
 Copyright 2005, R. Wayne Gallops

Dedication
There are numerous friends, mentors, and colleagues who have contributed to my
musical and academic growth. I am indebted to countless music teachers and students of
the Hillsborough County Public Schools, the University of Tampa, Hillsborough
Community College, Florida State University, the University of South Florida, and
Radford University who have offered encouragement and support. In particular, I thank
Jeanine Romano and Dr. Justin Askins for their statistical and editorial assistance. I am
also grateful to the late Dr. Virginia Bridges, who provided an opportunity to pursue
doctoral studies in music education through a fellowship offered in her name. Without the
creative vision of Dr. John Richmond and the diligent guidance of Dr. Jack Heller, Dr.
Sheila Woodward, Dr. David Williams, and Dr. Michael Robinson, this study would not
exist.
I am blessed with wonderful family members who have on many occasions,
provided moral and financial support for my academic endeavors. In particular, my wife
Donna, mother Janice, and grandmother Annie Lois Lee. I wish to dedicate this
dissertation to the memory of my late grandmother Annie Lois Lee, in honor of her
generosity and love of music.

TABLE OF CONTENTS
LIST OF TABLES iii
LIST OF FIGURES v
ABSTRACT vi
CHAPTER I: Introduction 1
Perception of Musical Expression and Stylistic Interpretation 4
CHAPTER II: Review of Literature 12

The Perception Contract 12
Conductor Effect on Musician Expression 16
Conducting Gesture Studies 20
Assessing Conducting Gesture 25
CHAPTER III: Research Procedures 29

The Research Question 29
Sampling Procedures 31
Procedure 32
Data Analysis 34
Measurement 36
CHAPTER IV: Data Analysis 38

Introduction 38
Process 38
Participants 39
Data Collection 41
Panel Evaluation 44
Inter-rater Reliability 49
Analysis of Data 50
Other Observations 65
Summary 67
i
CHAPTER V: Discussion 68
Educational Impact 71
Future Research 73
Synthesis of Conducting Domains 75
REFERENCES 77
APPENDICES
Appendix A: Demographics of Conducting Participants i
Appendix B: Musician Nesting With Conductor Videos ii
Appendix C: Unmarked Musical Excerpt iii
Appendix D: Marked Versions A & B of the excerpt iv
Appendix E: Sample of Marked and Evaluated Excerpts A & B
and Description v
Appendix F: Musician Response Scores for each Conductor Nest vi
Appendix G: Evaluation of Post (W/Conductor) Musician Responses ix
ABOUT THE AUTHOR End Page
ii
LIST OF TABLES
Table 1 Mean performance ratings and standard deviations

by body type and evaluator gender 18
Table 2 Mean scores of instructional conditions and dynamic

instruction by group 20
Table 3 Conducting participant demographics 40
Table 4 Conductor video performance evaluations

of the four-member panel 46
Table 5 Evaluations of musician baseline performances 47
Table 6 Musician mean response scores for each conductor 48
Table 7 GRI total scores 49
Table 8 Conductor Ranking: categorized based on GRI gesture score 51
Table 9 Conductor level and GRI response scores 54
Table 10 Conductor level and GRI response scores for each version (A&B) 54
Table 11 GRI Musician Response: ANOVA Procedure for conductor level

and excerpt 55
Table 12 Group means for conductor level and baseline-post

performances for each excerpt version 58
Table 13 ANOVA procedure for baseline-with conductor, conductor level,

and excerpt 58
Table 14 Baseline and Post GRI response scores for high and low level
conductors 59
iii
Table 15 ANOVA procedure for baseline-with conductor, conductor level,
and excerpt 60
Table 16 GRI Gesture: ANOVA procedure for conductor level and

excerpt (A & B) 63
Table 17 GRI Total: rank of conductors high to low 64
Table 18 Means for instrument type and response per excerpt 64
Table 19 ANOVA procedure for instrument type 65
iv
LIST OF FIGURES
Figure 1. GRI scoring box 27
Figure 2. Nesting procedure 34
Figure 3. Creation of the GRI Scores 36
Figure 4. Melodic excerpt (unmarked) 42
Figure 5. The Evaluation Process: sample of a highlighted copy

of excerpt A 45
Figure 6. Conductor and excerpt interaction for musician GRI response

Score 56
Figure 7. GRI Response score of high and low ability conductors with
excerpt 57
Figure 8. Excerpt A: musician baseline and post performance-responses 61
Figure 9. Excerpt B: musician baseline and post performance-responses 62
v
The Effect of Conducting Gesture
on Expressive-Interpretive
ABSTRACT
The purpose of this study was to examine the effect of non-verbal conducting
gesture on musicians’ stylistic response and whether conducting gestures alone elicit
consistent musical responses from musicians. Through an analysis that utilized a Gestural
Response Instrument (GRI) it was determined that, even if the use of verbal and facial
cues were eliminated, some experienced conductors successfully utilized non-verbal
conducting gestures to communicate specific musical interpretations. It appeared that
musicians responded in specific ways to the musical interpretation of conductors who had
command of a variety of conducting gestures. The results illustrated the existence of a
perceptual contract that facilitates the non-verbal communication expressed through
gestural conducting. As demonstrated through this study, some experienced conductors
lacked the gestural technique and vocabulary necessary to convey prescribed musical
decisions while others were proficient in this area.
vi
The Effect of Gesture
CHAPTER I
INTRODUCTION
This study examines how conductors utilize non-verbal gesture to communicate
stylistic interpretation that extends beyond accurate performance of notes and rhythms.
Many factors influence a conductor’s ability to communicate stylistic interpretation and
the musician’s response to the conductor’s communication. Musicians respond to the
musical interpretation of conductors beyond precision of notes and rhythms through a
process that involves nonverbal conducting gestures. This musical response to nonverbal
communication may illustrate a type of social-cultural contract among musicians.
Modern conducting technique, including the use of a baton, is a comparatively
recent musical development. Musical compositions of the late nineteenth century and the
continued increase in complexity and size of symphonic scores created an evolution of
modern conducting technique from mere time-beating to a highly complex and diverse
craft. “The problems that the contemporary conductor has to face are very different from
those of, say, fifty years ago...[there is] general demand for a degree of clarity, precision,
and smoothness of orchestral performance undreamed of even as recently as Richard
Wagner’s time" (Szell in Rudolf, 1950, p.vii).
1
The baton in its present form is the end result of hundreds of years of
conductorial experimentation. The earliest conducting was done with
gestures of the hand alone, describing melodic contour and pitches,
lengths of notes and phrases. Passing through stages where the leader sat
at the organ with his figured bass part and directed [therefrom]; where the
beat was an audible [prone], thumping along with the music; where the
concertmaster of the orchestra stood up and directed with his bow, and
finally where the silent beat with the baton arrived, conducting has now
become, through the use of the stick and its skilled sign-language, an art of
ultimate refinement (Green, 1961, p.7).
Three important domains that contribute to effective conducting can be identified:
1) knowledge and conceptualization of the score, 2) pedagogical ability to coach toward a
common interpretation, and 3) communication of interpretation through verbal and non-
verbal means (Wilcox, 2003). All three of these domains are perceived through the
mutual understanding of a common musical perception framework, a “social-cultural
contract” (Heller & Campbell, 1981). The historical and declarative elements of the score
contribute to a musical framework of informed intuition. Pedagogical knowledge and the
ability to perceive and shape musical sounds, based on a preconceived notion of the
score, contribute to conducting effectiveness. The conductor’s ability to communicate his
or her desires through a combination of verbal and nonverbal language completes the
conductor’s charge. Strength in all three domains should be the goal of a fine conductor.
Rarely is this goal achieved. The historical and theoretical requirements of score study,
the technical knowledge required of pedagogy, the personality aspects of verbal
communication, and the physical ability to communicate musical meaning through a
vocabulary of gestures are the elements that blend together on the conductor’s podium.
2
Imagine a clarinetist presenting a lecture/clinic at a professional meeting. After a
lengthy and rigorous discussion of the Mozart Clarinet Concerto K.622, the individual
begins to perform the piece. Given the performer’s vast knowledge of the piece, the
audience anticipates an extraordinary and beautiful performance. Instead, they are treated
to squeaks and squawks indicative of a beginning clarinetist. This clarinetist probably has
not developed the technical and procedural abilities necessary for communicating a
professional musical interpretation.
How typical it is that a performer “knows” much “about” a musical work but
cannot perform it very well. How often are fine musicians subjected to the “frumps” and
“flaps” of an extremely knowledgeable, yet gesturally deficient conductor? Many
suggest that because there are so many “gesturally deficient” conductors, musicians tend
not to look toward the podium (Berlioz, 1970). Some have implied that conducting
gestures have little or no meaning, particularly for professional musicians.
As it is, musicians tend to declare themselves conductors by merely

announcing as much to the world and ironically, sadly, the world generally
accepts the pronouncement without question, regardless of whether the
particular individual has conducting talents, regardless of whether the
individual has technical, intellectual, and emotional capacities to translate
a musical score into an appropriate interpretation via appropriate
conductorial gestures (Schuller, 1997, p.5-6).
An opposing view, as disseminated by many conducting pedagogues, is that a
well-rounded conductor, who possesses a non-verbal gestural vocabulary that is
3
consistent with an underlying agreement among musicians, is able to energize and bring
to the ensemble performance a heightened and consistent interpretation.
Noted conducting teacher, Elizabeth Green, suggests that when a “conductor's
impulse of will is strong and the technique is secure, then the ensemble truly has a leader
who can unify the musicianship of all into one secure interpretation. Such a conductor
has the finely developed technical skill and the confident drive to convey by his gestures
exactly what he wants without confusion or misunderstanding on the part of the players”
(1961, p.60).
Absent the domains of score study, pedagogy, and the sub-domain of verbal
communication, one can conclude that it is the ability to communicate through nonverbal
gesture that separates the conductor from the music historian, theorist, instrumentalist,
vocalist, or teacher. This study will focus on the effect of non-verbal conducting gestures,
absent facial cues, on the interpretive responses of instrumental performers.
Perception of Musical Expression and Stylistic Interpretation
The common ground that the performer (musician and conductor) and listener
have in terms of perception is the social-cultural contract. This social-cultural contract is
the musical environment in which one is raised. A common perception of musical
sounds within this musical environment is the foundation upon which conductors make
interpretive decisions. It is also the foundation from which conductors can effectively
communicate with other musicians in terms of stylistic interpretation. Absent this
4
common ground, transmission of musical ideas and the translation of those ideas into
gesture might be impossible. A discussion of conductor ability to convey expressive-
interpretive elements should be framed within an examination of this perception contract.
According to Dowling in Tighe (1993), musical understanding is a procedural
process developed through the cultural-historical context of one’s society. That is to say,
we learn about and come to understand music through acts of listening and performing
(Elliott, 1995). The social-cultural contract determines comprehension of musical
patterns through a set of implicit rules (Heller and Campbell, 1981). These implicit rules
are socially and culturally based, forming a common surface organization. The implicit
rules enable performers and listeners to organize musical sounds in a meaningful way and
collectively perceive nuance and interpret style. “Comprehension of different musical
languages and their respective nuances require listeners who are equipped with different
sets of implicit rules for recognizing and dealing with the surface structure features and
cultural performance conventions unique to each of those languages” (Fiske, 1990, p.6).
Through the social-cultural contract, musicians are able to evoke a musical style. If the
listener is also steeped in the same social-cultural milieu, then perception and musical
style can be realized as nearly the same. The implicit or procedural nature through which
we perform and perceive music is elaborated in the work of Heller and Campbell :
Perceptual learning in music probably involves an ability (similar to that

of speech) to detect those specific properties, patterns, and distinctive
features of the musical environment that are necessary for
communication. As in language perception, the rules necessary for
success of this active content dependent perceptual process are
implicit...and form a social/cultural contract between performer and
5
listener [that provides the context in which the acoustic signal is decoded]
(1981, p.21).
If the understanding and perception of music is so strongly based in an
internalized (procedural) social-cultural contract, what is the role of score study or
theoretical analysis in the preparation of musical interpretation? Perhaps the general
principles of form, harmonic structure, etc., that consume theoretical analysis are not key,
or as key, as one might think. For Leonard Meyer, “it is style and the musician’s grasp of
style that determines the framework” (1974, p.30). Perhaps it is the experiential knowing
through interaction with musical sounds that develops expression and stylistic intuition in
the musician. The artist “employs the rules and regularities of a prevalent paradigm -- the
grammar, syntax, and formal procedures of an existing style -- in order to create an
original pattern of particulars: a work of art” (Meyer, 1974, p13).
Understanding music is a process that may not be constrained by external (non-
musical) factors. According to Meyer, “the artists need not have conceptualized the rules
he uses let alone be able to explain them. His ‘knowledge’, based on practice and
experience, may be tacit” (1974, p13). Style is therefore predicated through intuition that
is developed within, and is the end product of the social-cultural-historical experience.
Information about form, dynamic contrast, tonal and harmonic relationships, and the
correct combination of pitch and rhythm are deciphered from the musical score.
However, these are not the complete elements of style in performance. How crucial is it
6
that performers have command of such “declarative knowledge” (Dowling in Tighe,
1993) of the score?
But were explanation a prerequisite for understanding, Bach might not

have been able to compose or Shakespeare to write. For, while both
supremely understood what they were doing, they might not have been
able to explain it. In short, composers, writers, and artists are especially
sensitive, skillful, and imaginative understanders of tonal, verbal, and
visual relationships. But much of their knowledge is tacit rather than
conceptual (Meyer, 1974, p.29).
Historical information might offer helpful anecdotes that contribute to the
performer’s overall knowledge of a particular musical composition. This information has
little to do with style, other than recognizing an already defined style period in which the
composition was constructed.
Style is the artist’s point of departure. It guides, but does not determine the
choices he [the composer] makes. Styles are complex probability
“systems” within which the artist chooses among alternative possibilities -
- that is, in terms of which he invents and combines patterns, develops and
resolves tensions, devises and elaborates formal relationships. For if the
rules specified precisely what was to be done, all choices would be
preordained, and originality would be impossible (Meyer, 1974, p.13).
The question begs, does the score specify precisely what is to be done? Through
analysis, musicians are able to develop declarative (theoretical) knowledge about music.
While Meyer and others assert that this knowledge has some, but not complete impact on
stylistic interpretation, developing declarative knowledge does contribute to refinement
of one’s procedural knowledge. Declarative knowledge develops informed intuition. This
7
informed intuition enables the musician to perceive complex relationships and elements
within the score and in musical sound.
Theoretical efforts to explain music tend to define the general aspects of
compositional structure and style. “Three kinds of hypotheses, used to explain works of
art, may be distinguished: (a) general laws, which are presumed to be constant over time
and space; (b) restricted principles, derived from and applicable to the norms and
procedures of a specific style; and (c) ad hoc reasons, which are necessary adjuncts to the
first two types when particular works of art are being explained” (Meyer, 1974, p.32).
So, there are general aspects to style that can be readily identified through score study.
Those aspects of the musical score constitute the dimensions that combine to form an
overall impression of the piece, but not the detailed stylistic interpretative dimensions
that create exceptional performance. This general, or macro dimension of the score, does
not in itself propagate the musician / conductor’s stylistic interpretation.
There must be a micro aspect of style that is not found in the score, but rather
through procedural knowledge, the social-cultural contract, and acoustic sound itself. It
is through the micro aspects of the score that conductors seek personal interpretation,
converting those musical notions into physical gesture.
Style is not only manipulated in the patterns that lie on the very surface (macro)
but at a deeper level (micro). “The formulations of style analysis are to general principles
as the taxonomy of species is to an hypothesis about their evolution or a theory of their
physiology” (Meyer, 1974, p.13).
8
Once a musical work is placed in its general taxonomy, the issues quickly become
micro. Therefore, musicians manipulate the acoustic sounds themselves in order to create
an interpretation that has been formulated through informed intuition. Decisions
musicians make in terms of shaping notes and fragments of phrases create stylistic
interpretations that can be perceived by audiences of similar cultural - stylistic
experiences. Lewis agrees with Meyer’s ideas on how musical style is realized. She
points out that:
A variety of detailed nuance in a musical composition is provided by

looking to the internal sublevels of interpretations and keeping them in
relationship with the whole. These internal sublevels of interpretation are
the minuscule levels of detail inside the phrase -- in the measure. They are
the tiniest levels of stress and restraint: acceleration and hesitation; weight
given a particular entrance or a final consonant; dynamic contour given
the little melodic motif, rhythmic twist, or harmonic detail; and attention
given to salient characteristics (Lewis, 1999, p.79).
The levels of detail discussed in Lewis’ (1999) work (stress, restraint, weight,
contour, etc.), actualize through the manipulation of note entrance, body, and release. The
musicians’ and listeners’ ability to perceive subtle inflections, and denote similar
meanings to those inflections, may be the essence of style.
The perception of microstructures, in the minutest detail, is verified through
Campbell and Heller’s work. Campbell and Heller (1979) demonstrate that the human
mind can quickly classify the timbral elements of sound. They (Campbell and Heller)
utilized this concept of microstructures in assessing listener identification of musical
instruments (timbre). “The traditional segmentation of a musical tone consists of the
9
attack transient, the steady state [body], and the decay transient [release]. We [Campbell
and Heller] have found it necessary to add the legato transient to this list, so that the
distinct character of the juncture in the two note sequence...can be indicated ” (1979,
p.21). How those fundamental variables of musical sound are manipulated and perceived
is the embodiment of stylistic interpretation. The challenge for the conductor is to
communicate how he or she wishes to manipulate subtle changes in these fundamental
variables.
The communication of musical expression from conductor to performer may be
reliant on the mutual perception of musical meaning as relayed through tension, release,
weight, and contour of note entrance, body, release and legato transient (between notes).
Conductors translate these perceived musical meanings into visual representations that
are, in turn, perceived by performers through an implicit mutual social agreement. This
social agreement may be referred to as a musician-conductor perception contract.
Through this performance agreement between conductors and musicians, a common
perception of musical meaning (style) and visual representation through conducting
gesture consistently manifests itself among trained musicians. The agreement of
conducting gesture may exist at a higher (or lesser) degree than commonly
acknowledged. Conducting pedagogue Elizabeth Green applied the term psychological
conducting to the idea that conducting involves a “transfer of ideas from the conductor’s
mind to the performer’s mind through the medium of correct and precise conductorial
10
technique without the use of verbal directions or written notation” (1961, p.241).
Through an experimental design, this study will investigate conductor ability to
communicate stylistic interpretation through conducting gestures alone (psychological
conducting) and examine possible commonalities that contribute to effective conducting
gesture.
11
CHAPTER II
REVIEW OF LITERATURE
The transformation of musical meaning into visual representation is the purpose
of the effective conductor. “A simple definition of the art of conducting could be that it
involves eliciting from the orchestra [any ensemble] with the most appropriate minimum
of conductorial (if you will, choreographic) gestures a maximum of accurate acoustical
results” (Schuller, 1997, p.9). This transformation of musical meaning into visual
representation may hinge on the existence of a perception contract that develops out of
the social-cultural contract of the listener. Research has shown that through musical
training, musical / acoustical perception is increased, thereby enabling the development
of a perception contract among musicians (Burnsed, 2001; Dowling, 1993; Fiske, 1990;
Fredrickson, 1997; Holahan & Saunders, 1997).
The Perception Contract
The existence of commonalities in listener perception of the interpretive aspects
of musical performance has been examined through numerous studies. The
communication of interpretive and stylistic decisions through conducting technique may
be reliant on a common music perception framework. Frederickson (1997) explored
questions of music perception in an experiment involving elementary, middle, and high
12
school students. In particular, do younger and older populations respond similarly to
perceived tension in music? Participants in the study were 2nd, 5th, 8th, 11th and 12th
graders. Participants from grades two and five were selected randomly from intact classes
at a suburban elementary school in a Midwest metropolitan area. Grade eight participants
were randomly selected from a middle school choir of the same metro area. Those from
grades 11 and 12 were randomly selected from a high school band of the same metro
area. The Continuous Response Digital Interface (CRDI), a device for measuring
electromotive forces, mounted on a dial and interfaced with a computer, was used to
collect data. Participants were asked to turn the dial to the right, as they perceived an
increase in degrees of tension in the music listening example.
Mean group responses were calculated for each grade level, once per second.
Once these data were calculated, Pearson correlations were calculated to compare the age
groups to each other. The data from this study were also compared to data of previous
studies that examined perceived tension in music and utilized the same procedures with
adult listeners. The results showed high positive correlations between groups of all age
levels as they responded to tension in music (r =. 86 - .98). When these results were
compared with a previous study (Madsen & Fredrickson, 1993) using adult subjects
(utilizing the same methods and procedures), high correlations were observed. The lowest
correlation was between the 2nd graders and adult musicians (r =. 71, p≤ .001).
13
While Fredrickson's study does not identify specific variables that elicit
perceptions of tension in music such as dynamic change, harmonic structure, dissonance,
etc., it is interesting to note that even young children have the capability of perceiving
interpretive variables beyond notes and rhythm, and that their ability to do so becomes
more profound with age and musical training. One must examine the results of these
studies with caution. Without reporting the variability of subject responses, the CRDI
Means could be misleading. The standard deviations of scores as well as the means
should always be reported. Without the standard deviations, correlations might give the
impression of much higher relationships than warranted.
Experiments reported by Dowling (1993) focused on ability to perceive tonal
transposition and preserve interval patterns in semitones. Musically experienced (older)
and musically inexperienced (younger) participants were selected and results of
their responses to listening patterns were recorded and compared to age/ability through
ANCOVA. The participants’ ability to identify interval patterns and tones as recorded
tone rows were altered was measured as the dependent variable. It was concluded that
five years of music lessons, at an early age, was sufficient to increase listening ability
in terms of identifying tonality and tonal-scale values. The “Moderately” and
“Professionally” trained group did not display enough difference to be statistically
significant. Dowling’s results suggest that music educators should emphasize procedural
learning so that students can gain acquired listening skills in the domain functions needed
for music cognition.
14
If a perception contract exists among listeners who hold a similar social-cultural
contract, musicians who grow from within the same contract may develop a refined
perception of stylistic interpretations. If musical training further develops the music
perception contract, conductors might gain access to that perception contract through
targeted methods. As musicians increase their ability to perceive interpretive variables,
conductors may gain stronger influence over the interpretive performance of musicians.
Common interpretive decisions communicated through common gestures that have
meaning to conductors and musicians might be instilled through the music training
process. Dowling’s study (1993) suggests that the ability to interpret subtle expressive
changes in music seems to be enhanced with musical training.
Like a phonetic alphabet, common listening skills enable the conductor to elicit
consistent and subtle responses from musicians. Although fundamental commonalities of
music perception can be demonstrated, Burnsed (2001) also found that it is developed
with age and musical training. The purpose of Burnsed’s study was to determine the
effects of expressive variation in dynamics on the musical preferences of middle school
and elementary school students. The populations from which subjects were sampled
included elementary students, middle school students, and music teachers. Two groups
were created and compared in this study. Group one included 288 students randomly
selected from grades 1-5 in an elementary school in Virginia. Group two
included 78 middle school music students and 22 choral and band directors participating
in a summer music camp.
15
Data reflecting subject preference for dynamic versus non-dynamic nuance were
reported through reliability coefficients (alpha) and Chi-Square preference significance
tests. These analyses showed that elementary students did not perceive a difference
between subtle expressions and non-expressive versions. The ninth grade students were
much more consistent in their preferences, as were the music teachers. While
specific musical variables were not identified, the results suggest that age and musical
training have impact an on perception of subtle expression in music. It could be,
however, that children may be able to discriminate, but they cannot label or effectively
use the CRDI.
Burnsed’s (2001) study, and the work of Fredrickson (1997), and Dowling (1993)
relate to the degree that conductors are able to control the expressive elements of a
musician’s performance and the degree to which listeners perceive elements of
expression in music performance. A technically skilled conductor may present a fine
visual representation of a musical passage, but if the musician performing the passage has
not developed an ability to perceive and perform subtle expression, the interpretation, as
communicated by the conductor, may not transpire.
Conductor Effect on Musician Expression
VanWeelden (2002) investigated the impact of conductor personal appearance on
the perceived quality of ensemble performance. Undergraduate music majors (N = 163)
were asked to complete a conductor effectiveness questionnaire after viewing a series of
16
videotaped conductor performances. Sixty-eight of the music major participants
(evaluators) were choral majors and ninety-eight were instrumental majors. The
conductor effectiveness questionnaire used a Likert type scale (1-5 point response) and
asked subjects to evaluate aspects of conductor effectiveness through eye contact, facial
expression, posture and overall impression. Videotapes of six female conductors
representing two different body types, ectomorphic /thin and endomorphic/large,
functioned as the independent variable. These conductors were videotaped in various
performance venues. Each conductor presented her best effort at conducting a
professional recording of a vocal work (The Coolin, by Samuel Barber). An analysis of
variance (ANOVA) of the participant responses determined that there was no significant
difference between endomorphic and ectomorphic conductors as judged by the students.
Table 1 displays the mean performance ratings and standard deviations by body type
found in VanWeelden’s study. These performance ratings reflect scores from the seven
questions that pertain to ensemble conducting effectiveness. Each male and female
evaluator produced scores within the range of 7 to 35 points for each endomorphic or
ectomorphic conductor
17
Table 1
Mean Performance Ratings and Standard Deviations by Body Type and Evaluator
Gender
Endomorphic Conductors Ectomorphic Conductors
Mean SD Mean SD
Male evaluators 24.12 3.08 24.29 3.36
Female evaluators 23.64 2.51 24.51 2.76
These findings suggest that musicians do not perceive conductor effectiveness based on
appearance, more specifically, body type. With this in mind, the assessment of conductor
communication and effectiveness should move beyond personal appearance and focus on
verbal and non-verbal behaviors.
Skadsem's (1997) study concerning the effect of conductor communication
addressed singers’ responses to verbal, written, and gestural-conducting stimuli. The
populations of interest for this study were choral conductors, college level singers, and
high school level singers. Ninety-six graduate and undergraduate music students and 48
high school music students from three metropolitan high schools surrounding a university
were pooled as participants. Data for the study were recorded on audiotape, videotape,
and on a written response questionnaire.
Singers were asked to perform a folk song (Michael Row Your Boat To Shore)
once with no conductor / written / audio cues, and eight times while viewing a conductor
18
and choir on a video monitor. Once an initial performance of the song (in its unmarked
version) was recorded, instructions regarding dynamic level were given for the remaining
eight performances. The dynamic level instructions were provided on the video screen
either through written cues, verbal commands, conducting gestures, or increase-decrease
in video audio volume (choir). In an effort to quantify the dynamic contrasts exhibited in
each performance, a panel of three expert conductors used the CRDI (Continuous
Response Digital Interface) to analyze audiotapes of individual vocal performances. An
average score for each excerpt performance was computed across the three judges. CRDI
data for the baseline (first performance with no cues) and the eight cued performances
were then compared in order to calculate a total change score. All of these data were
analyzed using an analysis of variance (ANOVA) and a post-hoc Tukey multiple
comparison test. These procedures of quantitative analysis enabled the researchers to
determine differences among groups and performances.
The data suggested that singers did respond differently to the four instructional
conditions that were presented: verbal commands, conducting gestures, written cues, and
increase-decrease of the choral audio. It was found that verbal commands made the
largest difference in soliciting dynamic response. Results suggested that singers respond
differently in loud and soft passages, more strongly reacting to instructions regarding
loud dynamics. The data also showed that the three groups (conductors, college students,
and high school students) responded differently to instructions about dynamics. The mean
scores of all instructional conditions by group are illustrated in Table 2.
19
Table 2
Mean Scores of Instructional Conditions and Dynamic Instruction by Group (Skadsem,

1997)
Condition Conductors College Singers HS Singers

Verbal soft 54.63 49.04 56.48
Verbal loud 51.38 47.38 50.67
Written soft 36.81 21.56 34.58
Written loud 18.54 12.17 20.48
Gesture soft 46.06 29.06 34.44
Gesture loud 39.23 21.10 4.06
Chorus soft 35.04 27.65 46.04
Chorus loud 14.04 24.83 28.71
The Skadsem study highlights the possibility that verbal communication may be
the most effective form of conductor communication. However, for more advanced
musicians, conducting gesture was a close second. The study suggests that ability to
communicate verbally is an important conducting skill and non-verbal behaviors,
specifically conducting gestures, also communicate musical ideas within the framework
of a musical social-cultural contract.
Conducting Gesture Studies
Nonverbal communication has been an area of psychological and educational
study for some time. Bulwer’s (1654) study of gestures used in public speaking is cited as
the earliest documentation of effective nonverbal communication. Numerous
psychological/educational studies affirm the need for effective nonverbal communication
skills in the classroom and music rehearsal. A more focused aspect of music education
20
research, as it relates to nonverbal communication, attempts to define conducting gesture
beyond verbal communication as nonverbal facial, body, hand and finger movement.
Efforts have been initiated to classify specific nonverbal conducting behaviors.
Sousa (1988) investigated the use of “emblems” in nonverbal conducting gesture. By
asking high school and junior high school students to describe video taped conductor
gestures, Sousa was able to label 38 out of 55 gestures as “emblems.” Sousa labeled a
specific gesture with emblem status if 70% of the subjects were able to effectively match
on a paper-and-pencil test the conducting gesture with its correct musical meaning. The
study determined that the level of gesture recognition increased with experience. High
school music students successfully recognized 37 of the 55 gestures while junior high
school students recognized only 19 of 55 gestures. The study found that in most cases,
musical experience increased ability to classify conducting gestures.
Cofer (1998) extended the work of Sousa (1988) by investigating the effect of
limited instruction in conducting gestural meaning on the musical response of seventh
grade band students. Participants were band students drawn from an urban public school
district (N = 60) who had two to three years of wind instrument performance experience.
Through random selection, an experimental group (n = 30) and a control group (n = 30)
were established. Participants in the experimental group received instruction on the
meanings of eighteen conducting gestures identified as emblems in the Sousa (1988)
study: piano, forte, subito piano, subito forte, two types of crescendo, two types of
decrescendo, staccato, legato, tenuto, two types of marcato, accelerando, ritardando, and
two types of fermata.

21
A four bar melody was constructed through which all of the musical expressions
could be demonstrated. Participants in the control group rehearsed the four bar melody
with an instructor who used no conducting gestures. The instructor of the control group
also reviewed with student’s terms that corresponded with specific musical expressions
identified for the study. During the preparation process, the instructor of the control
group tapped his baton on a music stand while students attempted to perform the melody
using the prescribed musical expressions. Participants of the treatment group were
instructed in the terminology of the musical expressions and corresponding conducting
emblems were demonstrated.
Conducting-gesture recognition of both groups (treatment and control) was
assessed through individual performance of the melodies while simultaneously following
the baton of a conductor on a television monitor, and a paper-and-pencil test. Results of
an independent sample t-test showed a significant difference between groups on the
paper-and-pencil measure of conducting gesture recognition (t = 6.97,p ≤ .001) and the
individual performance measure (F = 39.26, p ≤ .0001). The mean difference between the
treatment group (M = 16.40, SD = 2.68) and the control group (M = 12.50, SD = 2.68) on
the paper-and-pencil test reflects this result. The Bonferroni method, a statistical
adjustment that effectively raises the standard of proof needed when examining multiple
hypotheses simultaneously, verified that subjects in the treatment group performed nine
of the eighteen conducting gestures significantly better than subjects in the control group
(p < .05). Based on these results, Cofer concluded that seventh-grade band students of the
22
treatment group were able to consistently respond to nine of eighteen conducting
“emblems” after short-term conducting-gesture instruction while those of the control
group were only able to respond to three of the emblems. It is interesting to note that even
without instruction, the control group participants were intuitively able to correctly
respond to three of the 18 emblems.
Sidoti (1991) also explored the effect of expressive and non-expressive
conducting on the performance accuracy of high school students. Subjects in the
experimental group were asked to perform melodies while viewing a videotape of a
conductor utilizing expressive conducting gestures that visually reinforced expression
markings on the page. Subjects in the control group performed the same melodies while
watching a conductor who used non-expressive (beating time) conducting gestures.
Results indicated a significant difference between subject groups (p < .0001). Students in
the experimental group (M = 2.99, SD = .88) performed with greater accuracy, in terms
of selected expression markings, than did students of the control group (M = 2.78, SD =
.80). It is interesting to note that the Sidoti study utilized an upper body video of the
conductors with no isolation of facial or body language cues, and that participants
performed from copy that included expressive markings.
House (1998) approached the question of conductor influence on the expressive
performance of musicians in a similar way. College trumpet students were instructed to
perform an excerpt while following a conductor on video. A leading instrumental
conductor (John Whitwell) was video taped while conducting the excerpt in an expressive
23
way and once again in a non-expressive way. The expressive approach made an attempt
to convey the interpretive markings of the excerpt. In contrast, the non-expressive
conducting approach utilized simple beat patterns. Musicians performed the excerpt
while viewing the conductor videos. A three-member panel then evaluated the
expressiveness of those performances.
Even though the musicians performed from a copy of the excerpt which contained
expressive-articulative and phrase markings, a significant yet small difference in the
expressiveness between musician performances recorded while viewing the non-
expressive conducting (M = 4.896, SD = 1.606) versus the expressive conducting
(M = 5.287, SD = 1.732) was evident (p < .01).
The use of videotaped conducting as a two-dimensional medium for isolating
conducting gesture presentations has been implemented not only in the House study, but
also in many other studies (Cofer, 1996; Fredrickson, 1992; Sidoti, 1990; Skadsem, 1997;
Sousa, 1988). Even though responding to a live conductor is optimal in terms of
ecological validity, the use of videotaped conducting ensures that participants are
exposed to a consistent independent variable.
Graves (1984) isolated verbal communication and facial/body language cues from
technical conducting gestures. The Graves study applied six experimental conditions in a
choral rehearsal context. An unusual, and interesting, dependent measure was used in the
Graves study. The performances of a Bach chorale were recorded under the different
conducting conditions and then presented to high school students to determine if they
24
heard any differences in the performances. Conductors were asked to prepare the Bach
chorale using verbal and nonverbal conducting techniques, nonverbal techniques alone,
and nonverbal techniques while wearing a facial mask. Results indicated that there
were no significant differences between the nonverbal conditions and the nonverbal
/masked conditions. This outcome suggests that, while facial expression is a valuable
element of nonverbal communication, it is not a crucial variable on its own or when
removed from nonverbal conducting gesture.
Assessing Conducting Gesture
Herbert Karpicke’s study (1987) investigated conductors through the development
of an instrument that measures the performance response of musicians to a variety of
nonverbal conducting gestures. Karpicke defines conducting gesture as “an observable
movement of the hands or hands and arms, or a change in visual direction, in addition to
or outside of an ongoing beat pattern which results in observable musical response in
performance context”(1987, p.3). The purpose of his study was to develop a reliable and
valid instrument for evaluating and measuring conductor gesture and subsequent
performance (musician) response.
25
Two subject groups were selected for experiments that would validate Karpicke’s
Gestural Response Instrument (GRI). Twenty-five high school orchestra conductors from
the Houston and Dallas, Texas areas were contacted for participation. Twenty (n = 20)
conductors agreed to participate in the experiments. Although this was a convenience
sample, the researcher was able to obtain a great deal of demographic and performance
related information on these subjects. Each conductor had a record of participation in
Texas performance evaluation festivals. Their respective music supervisors verified this
information. Conducting subjects were placed into two categories, “strong” or “weak,”
based on previous festival ratings and music supervisor knowledge of the subject’s
conducting performance ability. The conductors were then rank ordered (strongest to
weakest) by their music supervisors.
Conducting subjects conducted a volunteer orchestra through a series of musical
excerpts. Prior to the experiment, the researcher rehearsed the piece that included these
excerpts.
The conductors were instructed to use no verbal cues or communication during
the performance/recording sessions. Only facial and hand/arm conducting gestures were
utilized. Participants conducted the piece while simultaneous video recording took place.
Excerpts were cued at consistent places in the music. Six of these excerpts were selected
from each conductor-orchestra performance.
Four expert observers were trained in using the GRI and were asked to view a
random mix of video excerpts from each conductor’s performance, totaling 560
26
observations. Using a 1-5 Likert scale, observers were asked to rate the perceived
effectiveness of each conductor gesture (top box of the GRI) and the corresponding
musical response of the orchestra (Figure1).
Figure 1. GRI scoring box
4
gesture
3
response
Gesture (top) and response (bottom) scores for each excerpt were multiplied and
averaged, creating an overall Gestural Response Score. These scores were then aligned
with their respective conductors. ANOVA (Analysis of Variance) was used to verify
consistency among conductors, observers, musical excerpts, gestures, responses, and
gesture/response variance.
According to Karpicke, this scoring method “allows both the gesture and its
associated response to be scored as contingencies” (1987, p.59). When four observers are
used, the GRI Total scores yield a substantial reliability (r =. 80). It was also noted that
observer consistency remained stable even though musical content (excerpts) varied.
Because the same orchestra performed the same excerpts with each conductor, it is
important to know which facet was having a stronger effect, the conductor or maturation
of the orchestra. In this study, no systemic orchestra response was found: however, it
27
would seem that consistency of musician response and learning effects (history,
maturation, practice effect, etc.) should be accounted for with use of the GRI in a group
performance setting.
Data suggest that with four or more observers, the GRI is considered a reliable
instrument for assessing conducting gesture/musician response events. The GRI is also
able to discriminate between stronger and weaker conductors' ability to elicit musical
responses through gesture. It should be noted that response score reliability (r =. 50) was
lower than gesture score reliability (r =. 85). This discrepancy suggests that observers
should receive further training in how to consistently assess the aural/response aspect of
the GRI instrument. Perhaps audio of the musical response should be evaluated
separately rather than simultaneously with video of the corresponding gesture.
The use of a full ensemble brings a number of confounding variables into
Karpicke’s study. Certain variables may be controlled (history, maturation, bias, ability,
following the principal player, etc.) using individual musicians and recording individual
responses.
Karpicke’s Gestural Response Instrument (GRI) study is a substantive effort to
develop a subjective assessment of conductor performance. Use of the GRI and other
methods of conductor assessment may enhance conductor training and examine musician
perception of conducting gestures.
28
CHAPTER III
RESEARCH PROCEDURES
The Research Question

For musicians, it is widely believed that the ability to understand and manipulate
segments of musical sound, such as entrance, body, and release, are crucial for musical
expression (Lewis, 1999; Gallops, 1999). Stylistic interpretive decisions are made
through the musician’s informed intuition of appropriate musical shape and ability to
“control the beginning, middle, and end of notes” (Gallops, 1999, p.35). For conductors,
perception and manipulation of these segments of musical tone (entrance, body, release)
may be crucial when translating musical interpretation into visual representation. Elliott
articulates the procedural process of informed intuition involved in music making as an
action involving “nonverbal forms of thinking and knowing in and of themselves”
(Elliott, 1995, p.55). As discussed in previous chapters, stylistic expressive-interpretive
decisions are generated through the conductor’s informed intuition of appropriate musical
shape.
The ability to translate musical interpretation into visual representations, the
perception and ability to manipulate entrance-body-release, and the informed intuition
developed through experience and knowledge, are tools that engage conductors in a
perception contract. This perception contract among musicians may allow nonverbal
communication of very specific interpretive decisions. According to this view, the
conductor’s intuition and learned ability to communicate those stylistic-interpretive
decisions through gesture is what ultimately determines conducting success (Graves,
1984; Grechesky, 1985; Green, 1961; Lewis, 1999; Mayne, 1992; Miller, 1988;
29
Strouse, 1987). In order to establish the existence of a perception contract between

musicians and conductors, this study investigated differences between performances of
musical excerpts (baseline-post w/conductor) through an experimental design using a
Gestural Response Instrument - GRI (Karpicke, 1987) as the dependent measure.
In an effort to minimize confounding and extraneous variables, a Gestural
Response Instrument (GRI) format was used in this study with three modifications; 1)
individual musicians performed the musical excerpts rather than a full orchestra, 2) each
conductor’s performance was video taped, focusing only on the hands/arms, 3) panels of
experts evaluated the conducting videos separately from the audio record of musician
responses, and 4) a nested sampling procedure was utilized in an effort to increase
observation responses while minimizing order effect and practice effect.
Descriptive data were collected and examined. Effect sizes between performances
demonstrated the degree to which communication occurred through nonverbal
conducting gestures. Demographic information pertaining to each conductor was
gathered. This information included gender, age, conducting background (band or
orchestra), experience, type of training, major instrument, and level of professional
involvement (high school, college, university, and professional). After data analysis,
trends appeared in individual conductors. These trends will be discussed in Chapter IV
and Chapter V, and may warrant further research.
Instrumentalists gather their musical experiences from varied ensemble traditions.
Thus the string musician might be more responsive to expressive-interpretive conducting
than the brass player. Moreover, the wind player might create clearer interpretive
articulations. If this is the case, differences in brass, woodwind, and string musician
response scores may appear. The sample size of brass, woodwind, or string
30
musicians for this experiment was not large enough to generalize significance from an
ANOVA in nested sampling (Kromrey & Dickenson, 1996). However, results of this
study might display a trend that warrants further research with a larger generalizable
sample.
Sampling Procedures
Two pools of participants were sampled for this study: experienced instrumental
conductors and college musicians who performed musical excerpts while watching the
conductors. Conductors had equal opportunity to be selected from a pool of 50
experienced high school, college, university, and professional instrumental conductors.
Through simple random sampling, fifteen conductors were selected from the pool of 50.
This study focused on experienced instrumental conductors, particularly those who have
demonstrated success as a high school, college, or professional conductor. In order to
examine trends for future research and to clarify results as related to individual
conductors, preliminary information was gathered from each conductor that included
gender, specific age, conductor background (band or orchestral conducting), type of
conducting experiences, training, major instrument, and current area of professional
involvement: high school, college, university or professional (see Appendix A).
Woodwind, brass, and string instrumentalists were selected from a pool of fifty
college/university level music majors. A music major was defined as someone who is
pursuing an undergraduate course of study in music performance, music education, or
composition at a college or university. Twenty-five participants were selected through
stratified random sampling, providing numbers of woodwind-brass-string instrumentalists
(brass n=10 / woodwind n=10 / string n=5).
31
Procedure
One melodic excerpt was selected. The excerpt did not contain extremely
complex rhythms or technical challenges and the excerpt was not easily recognized from
standard classical literature. That is to say, it was not familiar to the performing
musicians as commonplace among instrumental performance literature or numerous
recordings. Once an appropriate excerpt was identified, it was duplicated three times.
One copy of the excerpt contained no expressive-interpretive markings. Distinct
expressive-interpretive markings were created for each of the two remaining copies of the
excerpt. These marked versions of the melodic excerpt were labeled as version ‘A’ and
version ‘B’ respectively. An expressive-interpretive marking was defined as staccato,
tenuto, accent, slur, dynamic, crescendo, decrescendo, accelerando, ritardando, and
tempo. (Appendix D)
Conductors selected for participation practiced and conducted each of the marked
excerpt versions (‘A’ & ‘B’). In preparation, conductors used whatever amount of time
they needed to prepare and practice each version of the excerpt. Conducting
performances were video recorded while participants conducted, using gestures believed
to represent the expressive-interpretive markings of each excerpt. While recording the
gestures, conductors were allowed to sing as they conducted and were asked to use a
baton and both hands as they saw fit. Through use of digital video technology, the face
was blurred and any vocalization was removed from the recording. Only hands/upper
chest were visible on the video, thereby eliminating overt body language and facial or
vocal cues. Each excerpt was duplicated with a fifteen-second pause between repeated
conductor performances of the excerpt. All videos were coded and timed, identifying the
32
conductor and the excerpt (‘A’ & ‘B’) that was on the video. A separate VHS videotape
was generated for each conductor.
From the selected musicians (n=25), randomly selected nests of 5 participants
watched and performed individually to mixed sets of conducting videos. Each nest
included 2 brass, 2 woodwind, and 1 string musicians. Nests were pre-assigned to sets of
six video-conducting excerpts. This process was duplicated, thereby creating enough
nests to perform all of the video excerpts. Through the nesting procedure, an increased
number of responses was generated with limited duplication of personnel. Limiting the
number of times that a participant performed the excerpts, yet generating a large set of
responses was an important result of nesting. VanWeelden (2002) found that conductor
performance ratings were affected when participants viewed conductors in a certain
order. Suppression of practice effect and order effect in participant response was an
important design aspect of this study (Appendix B). The nesting procedure applied in this
study controlled for suppression of multiple response (practice effect) and created a
mixed order of videos (Dunham, Lemke & Moran, 1991; Holahan & Saunders, 1997;
VanWeelden, 2002). Care was taken in determining the unit of measure as the mean of
each nest. Some research has shown that power is lost when nested sampling is applied,
particularly when individual scores are the unit of measure (Kromrey, 1996).
Participants received and practiced the unmarked copy of the excerpt, focusing
simply on accuracy of notes and rhythms. The musicians performed individually in
consistent acoustical environments, with no other participants present. Each musician was
instructed to expressively perform the melody using a personal interpretation. This initial
performance was labeled as that particular musician’s baseline response. The musicians
were then asked to follow the conductor’s video taped gestures as he/she performed each
33
excerpt. The musician had an opportunity to perform each excerpt twice while viewing
the conductor video (pausing fifteen seconds between each performance). These two
performances created a pre-post (with conductor) set of performance recordings for each
version of the excerpt (Figure 2). Figure 2 illustrates the mixed assignment of conductor
videos as they were assigned to randomly selected nests of musicians.
Figure 2. Nesting Procedure
Pre
A b1
B b5
B w6
A w10
A s8
11 B b6
15 A b2
A w2
A w5
B s6
b1= brass 1; w6= woodwind 6; s8=string 8
Data Analysis
Using the format of Karpicke’s Gestural Response Instrument (GRI), panels of
experts evaluated the conducting gesture effectiveness and the musician performance
responses (baseline and post w/conductor) as compared to the expression markings of the
34
‘A’ and ‘B’ versions of the excerpt. For the conductor evaluators, an expert was defined
as a musician with graduate training in instrumental conducting and professional
performance experience as a musician and conductor. For the musician response
evaluators, an expert was defined as a musician with undergraduate training in
instrumental music and professional performance experience. Each panelist received
training and practice in highlighting (marker pen) perceived expressive-interpretive
performance, using a five point Likert Scale, and utilizing the GRI gesture or response
scoring boxes.
The conductor evaluation panelists applied a Likert Scale to evaluate each
conductor video recording (N=15, 30 excerpts) in terms of gestural conducting
effectiveness (GRI Gesture). The Likert Scale categories were (5) definite gestures, (4)
somewhat definite gestures, (3) inconsistent gestures, (2) not very definite gestures, and
(1) poor gestures. The task of each panelist was to rate the conducting gestures as they
attempt to express through their performance of the phrase the expressive and
interpretive markings of the excerpt. These GRI ratings were assigned without listening
to any performance. The decisions of the panelists were based on how well the
conducting gestures seemed to represent the two written versions of the excerpt.
Audio recordings were collected as musicians performed the excerpts. The audio
recordings were labeled and sorted according to excerpt (‘A’ & ‘B’), baseline or post
performance (w/conductor), and conductor (N=15). When comparing the performances to
the marked excerpt, the musician response panelists evaluated the performances in terms
of expressive-interpretive accuracy. The panelists were asked to mark with a highlighter
pen while listening to the recordings, those interpretive markings that the musicians
adhered to in the performance. A five point Likert Scale was used to quantify each
35
panelist’s evaluation of the musician responses (GRI Response). The Likert Scale
categories for musician response were defined as (5) definite responses, (4) somewhat
definite responses, (3) inconsistent responses, (2) not very definite responses, and (1)
poor responses.
Measurement
Gesture and Response scores were observed individually and were multiplied to
create a GRI score for each performance (as recommended by Karpicke). GRI scores
were aligned with conductors and a total GRI score was calculated (averaged) for each of
the excerpts. A Conductor GRI score was calculated (averaged) from the excerpt scores
to give each conductor an overall GRI gesture, response, and total score (Figure 3).
Figure 3. Creation of the GRI Scores
GRI Gesture GRI Response GRI TOTALS overall totals

conductor excerpt score BASELINE POST BASELINE POST Baseline Post
(w/conductor) (w/conductor)
1 1a 4 w3 b3 s4 w4 b4 s5 4 / 3.33 4/4 13.33 16
1b 3 w3 b2 s3 w3 b3 s3 3 / 2.67 3/3 8.00 9
N=15 conductors w=woodwind b=brass s=string
Once these descriptive data were acquired, differences in GRI scores between
performances of the excerpts (baseline-post w/conductor) were examined. If non-verbal
communication took place, GRI response scores should increase each time a musician
36
performed an excerpt with a specific conductor. An ANOVA calculated differences in the

average GRI scores between baseline and post performances (w/conductor) of the
excerpts. Chapter IV displays those differences, concluding that strong and specific non-
verbal gestural communication can take place between conductor and performer.
37
CHAPTER IV
DATA ANALYSIS
Introduction
This chapter will illustrate and elucidate data that were collected from conductor
videos, musician responses, and the evaluation of those performances-responses. The
analysis will be guided through a discussion of process, participants, collection of data,
results, and reliability of the results.
Process
The Gestural Response Instrument (GRI), as described in the previous chapter,
was utilized in this study. The product of a gesture score and response score creates a
total score that reflects the ability of a conductor to communicate a musical interpretation
through conducting gesture, and the corresponding response of musicians to those
gestures. It is important to understand what each of these scores represents.
The GRI gesture score is the average score of a four-member panel evaluation of
the conductor as he/she conducts a passage that is marked with specific interpretive
markings. Evaluators rated each conductor’s ability to communicate the interpretive
markings of the passage with a 1-5 numerical Likert score. Two contrasting versions of
the excerpt were created in order to verify results were not obtained due to musician
intuition or innate properties of any particular expressive-interpretive approach. With this
38
in mind, two GRI gesture scores were calculated: one for an ‘A’ version of a melodic
excerpt and one for a ‘B’ version of the same excerpt. Those two scores were averaged,
creating a GRI gesture total.
The GRI response score is the average score of five musician performances
(responses). A three-member panel evaluated these performances. Each evaluator
assigned a Likert Score of 1-5 based on how closely the performance matched the
specific interpretive markings of the ‘A’ or ‘B’ version of the melodic excerpt.
The gesture score and response score for each conductor was multiplied, creating
a GRI total score. For example, if a conductor received an overall gesture score of 4 and
a musician response score of 3, the GRI total score for that conductor was 12. As
described in the Karpicke (1987) study, the GRI total score can therefore range from 1 to
25. If the two scores were added instead of multiplied, the total possible score would be
10. The multiplied score allows a greater range (25 instead of 10).
Participants
The participants involved in this study were categorized as conductors and
musicians. As discussed in the previous chapter, conductor participants were randomly
selected from a pool of experienced conductors. In order to determine the effect of
gesture on the performance of musicians, it was important to control for experience and
training of conductor, therefore all of the conductors represented successful, trained,
instrumental conductors. The conductors who actually participated in this study reflect
the following demographics (Table 3).
39
Table 3
Conducting Participant Demographics (N=15)
Number of Conductors out of N=15

GENDER
Female 3
Male 12
EXPERIENCE
High School Conducting 14
College-University Conducting 9
Community Ensemble Conducting 7
Professional Conducting 3
TRAINING
Undergraduate Degree in Music 15
Masters Degree in Music 9
Doctorate-Post Masters Degree or study 3
Numbers reflect how many of the conductors (N=15) fit into each category.
For detailed information on each conductor: see appendix A
The musicians who performed the melodic excerpts were college-university
undergraduate music majors representing brass, woodwind, and string instrumentalists
( N=10 brass / 10 woodwind / 5 string). The GRI response score assigned to each
conductor was an average of five musician response scores.
Musicians were assigned specific conductors and excerpts through a nesting
procedure. A nest of five musicians representing two brass, two woodwind, and one
string player viewed each conducting performance. The randomness of musician
assignment limited order effect and created 75 recordings for each version of the excerpt,
plus 25 baseline performances, totaling 175 recorded performances. For a detailed listing
of musician assignments, see Appendix B.
40
A four-member panel of expert evaluators rated the video taped performances of
each conductor. The conductor evaluators were expert musicians who have extensive
experience as conductors, professional performers, and university music professors.
Due to the large number of performance recordings generated, two three-member
panels evaluated the recorded responses of the musicians. Each panel was comprised of
experienced professional musicians. Panel ‘A’ rated the performances of the first half of
recordings and panel ‘B’ evaluated the remaining half.
Data Collection
As discussed in Chapter Three, this study utilized digital video and audio
technology throughout the data collection phase. A few days before they were to perform
for the camera, conductor participants were provided with two distinct versions of the
melodic excerpt selected for the study (Der Jungling Am Bache, by Schubert): version
‘A’ and version ‘B’ (Appendix D).
Each conductor was instructed to review and practice both versions of the melodic
excerpt, keeping in mind that the goal was to conduct in such a way to elicit an
interpretive response from a musician as close as possible to the interpretative markings
on the written excerpt. The conductor was asked to use a baton and was informed that
facial cues would be digitally removed from the video. A blank wall was used as a
background and the digital video camera was placed approximately 12-15 feet away. The
camera view was framed above the conductor’s waist so that all of the beat patterns and
41
gestures could be recorded. Conducting participants were also instructed that they could
sing, whistle, or hum as they conducted if they wished, and that the audio would be
turned off when the musicians viewed the video. A conductor code (1 to 15) and an
excerpt code (‘A’ or ‘B’) were verbally cued at the beginning of each videotaped
conducting performance. Once each conductor rendered performances of both versions of
the excerpt, a digital “video blur” was placed over the face of each conductor. The digital
blur prohibited musician participants from deciphering facial gestures that might relate to
the music. This process also inhibited the identification of the conducting participants.
The product was a master video of fifteen conductors conducting both versions of the
excerpt in a way that, at least in the mind of the conductor, communicated as much of the
interpretive details as possible.
Musician participants were recorded while performing from an unmarked copy of
the melodic excerpt (Figure 4). Although it retains rhythmic and melodic elements, the
unmarked copy of the excerpt is void of any interpretive-phrase markings.
Figure 4. Melodic Excerpt (unmarked)
42
The recording equipment consisted of a dual cassette recorder, two omni-
directional microphones, and a 20” television monitor with VCR playback. The
microphones were placed at the optimum distance appropriate for each instrument. Each
musician was given the option of performing while seated or standing with the television
monitor placed at eye level, two and one half feet away.
Musicians were given the unmarked copy of the excerpt in advance and were able
to practice the excerpt until it was well under hand. Once warmed-up and familiarized
with the recording surroundings, the musician was asked to perform the excerpt utilizing
his/her own personal interpretation. This initial, or baseline performance, was recorded to
provide a reference from which changes in future performances could be measured. The
baseline performance demonstrated a musician’s intuitive interpretation prior to the
conductor-initiated performances.
Once the baseline performance was recorded, the musician was instructed to
perform the excerpt while viewing six video recordings selected from the thirty conductor
video performances. The musician was asked to interpret the excerpt in the way that
he/she felt the conductor was communicating musical interpretation through conducting
gesture. The musician’s performance response was recorded on audio cassette tape and
labeled with the corresponding conductor-excerpt code. This process created a baseline
and six conductor initiated performances for each musician (N=25). A detailed listing of
musician codes and the corresponding nesting of conductor videos can be found in
Appendix B.
43
It is important to note that musicians performed along with each conducting video
twice with fifteen-second intervals between each; first as a practice recording (pre), and
secondly as a “post” (w/Conductor) recording. The post (w/Conductor) recording was
used as data for the evaluation procedure. The recording process was not halted, even in
those cases where the musician could not follow the conductor’s gestures.
Panel Evaluation
Panels of qualified professional musicians evaluated the video and audio
recordings produced by the conductors and musicians. The evaluation process utilized a
1-5 Likert scale scoring system and a procedure that involved highlighting interpretive-
articulative markings on copies of the melodic excerpt.
The conductor evaluation panel was instructed to view each conductor conduct
each version of the excerpt (A & B) while following along with the matching marked
version of the excerpt. As the conductor utilized patterns and gestures that the evaluator
regarded as effective communication of an interpretative mark on the score, the evaluator
was to highlight (with a marker pen) the specific interpretive-articulative markings
(Figure 5).
44
Figure 5. The Evaluation Process: Sample of a Highlighted Copy of Excerpt ‘A’
Once the evaluator viewed the conductor’s videotaped performances and
highlighted the interpretive-articulative markings that, in the view of the evaluator, were
effectively communicated, a 1-5 Likert score was assigned. As discussed in Chapter
Three, the Likert scoring system was defined in the following way: (5) definite gestures,
(4) somewhat definite gestures, (3) inconsistent gestures, (2) not very definite gestures,
(1) poor gestures. The task of the evaluator was to rate the conducting gestures as they
attempted to express the articulative and interpretive markings of the excerpt. As
illustrated in Table 4, this scoring process assigned four ratings to each conductor (N=15)
for each version of the excerpt (A & B). A mean score for each conductor was calculated
and termed as the GRI Gesture score.
45
Table 4
Conductor Video Performance Evaluations of the four-member panel: GRI Gesture Score
Evaluator (I-IV) – A or B version of the excerpt
Mean=
GRI Gesture
IA IB IIA IIB IIIA IIIB IVA IVB score
CONDUCTOR
1 3 3 3 3 4 3 4 3 3.25
2 2 2 2 2 1 1 3 3 2.00
3 4 4 2 2 4 4 4 4 3.50
4 2 2 4 4 4 3 3 3 3.12
5 4 5 3 3 3 3 3 3 3.37
6 3 3 3 2 3 1 1 1 2.12
7 2 2 1 2 3 2 4 4 2.50
8 3 3 2 2 3 2 3 3 2.62
9 2 1 2 2 2 1 2 2 1.75
10 4 4 3 3 4 4 2 3 3.37
11 2 2 2 2 3 2 3 2 2.25
12 3 2 4 3 3 2 2 2 2.62
13 2 2 4 3 3 3 3 4 3.00
14 1 1 1 1 2 1 2 2 1.37
15 5 4 5 4 4 4 4 5 4.37
A similar process was used to evaluate musician responses to the conducting
gestures. Due to the number of recordings created (175), two three-member panels of
musician evaluators listened to the musician performances. These evaluators were
instructed to track the performances with the corresponding version of the excerpt,
marking the interpretive-articulations that were perceived by the evaluator.
A 1-5 Likert score was assigned to the musician response (performance),
reflecting the evaluator’s perception of the degree to which the musician performed the
specific interpretive-articulative markings of the excerpt. The Likert Scale categories for
musician response were; (5) definite responses, (4) somewhat definite responses, (3)
inconsistent responses, (2) not very definite responses, (1) poor responses.
46
The Baseline performances created by each musician were scored against the A
and B versions of the excerpt. Since the baseline performance was the musician’s own
personal interpretation of the excerpt, and because the baseline performance was
compared to the ‘A’ and ‘B’ marked version of the excerpt, the evaluator’s Likert rating
was typically low (1.0 < 2.66). The Likert ratings for baseline performances as they were
evaluated against the ‘A’ and ‘B’ versions of the excerpt, are displayed in Table 5.
Table 5
Evaluations of Musician Baseline Performances

Musician Code Excerpt ‘A’ / ‘B’ Mean score
B1 (1) 1.33 / 1.33
B2 (2) 2.0 / 1.66
B3 (3) 1.66 / 1.66
B4 (4) 1.0 / 1.0
B5 (5) 1.66 / 1.33
B6 (6) 2.33 / 2.0
B7 (7) 1.66 / 1.66
B8 (8) 2.66 / 2.0
B9 (9) 2.33 / 2.66
B10 (10) 2.0 / 2.0
W1 (11) 2.0 / 2.0
W2 (12) 2.33 / 2.33
W3 (13) 2.33 / 1.33
W4 (14) 1.66 / 1.33
W5 (15) 3.0 / 3.0
W6 (16) 2.0 / 2.0
W7 (17) 2.0 / 2.0
W8 (18) 1.66 / 1.66
W9 (19) 2.0 / 2.33
W10 (20) 2.33 / 3.0
S1 (21) 1.66 / 1.66
S2 (22) 2.0 / 1.33
S3 (23) 2.33 / 2.0
S4 (24) 1.33 / 2.33
S5 (25) 2.0 / 2.0
B=Brass Musician Code / W=Woodwind Musician Code / S=String Musician Code
47
The musician response scores were aligned with each conductor. The mean for
the Post (w/Conductor) musician responses of each nest was calculated in order to create
a GRI Response score for each conductor. Table 6 displays the musician mean response
scores for each conductor’s nest of musicians.
Table 6
Musician Mean Response Scores for Each Conductor
CONDUCTOR GRI Response Score

1 3.15
2 2.53
3 2.87
4 2.97
5 3.20
6 2.66
7 2.60
8 2.77
9 2.23
10 3.4
11 2.43
12 2.97
13 2.70
14 2.23
15 3.43
The GRI Response score is the mean of the musician response score of the nest that view the conductor.
Table 7 reflects the GRI TOTAL scores for each conductor. This score is the product of
the GRI Response and GRI Gesture score. The GRI TOTAL score quantifies the effect of
the conductor’s gestures on the musician’s responses (Karpicke, 1987).
48
Table 7
GRI Total Scores
CONDUCTOR GRI Gesture Score GRI Response Score GRI TOTAL

(GRIG) (GRIR) (GRIGxGRIR)
1 3.25 3.15 10.24
2 2 2.53 5.07
3 3.5 2.87 10.03
4 3.12 2.97 9.27
5 3.37 3.20 10.8
6 2.12 2.66 5.65
7 2.15 2.60 6.5
8 2.62 2.77 7.26
9 1.75 2.23 3.91
10 3.37 3.40 11.47
11 2.25 2.43 5.47
12 2.62 2.97 7.78
13 3 2.70 8.1
14 1.37 2.23 3.07
15 4.37 3.43 15.02
Range = 3.07 < 15.02
Inter-rater Reliability
Threats to validity can arise in studies that involve subjective assessment if inter-
rater reliability is not estimated and accounted for. This study involved the evaluation of
conductors by one panel of raters, and the evaluation of musician responses by two other
panels of evaluators. It is important to remember that subjective rating, especially for
musical performance, will typically generate medium-high reliability estimates (r =.55 -
.80). It is important to examine reliability quotients as they relate to literature in the field.
This particular study utilizes the Gestural Response Instrument (GRI) developed in the
Karpicke study of 1987. Reliability quotients of the Karpicke study ranged from r =.50 to
r =.80.
49
The first examination of inter-rater reliability involved an evaluation of the
conductors as they conducted the two versions of the excerpt. The estimated reliability of
the panel when they evaluated all conductors on the ‘A’ version of the excerpt was
r =.38. The reliability estimate for the evaluations of the ‘B’ version was r =.55. The
reliability of the ‘A’ version (r =.38) evaluations is quite low. The ‘B’ evaluations fall
within the medium range of reliability (r =.55).
Since there was no significant difference found between versions of the excerpt
and conductor GRI Gesture scores, and given the results of the initial reliability estimates
for the GRI Gesture scores, the researcher extracted the Mid-ability conductors and
calculated reliability estimates for the High and Low conductor evaluations. The
estimated reliability of the panel when they evaluated High and Low ability conductors
on the ‘A’ version of the excerpt was r =.56. The reliability estimate for the evaluations
of High and Low conductors on the ‘B’ version was r =.68. Extracting the mid-ability
conductors increased the inter-rater reliability estimates.
Reliability was also measured between raters of the musician responses. Once
again, these estimates reflect the evaluation of responses to the ‘A’ version and the ‘B’
version of the etude. The reliability for the evaluations of musician responses for the ‘A’
version of the etude was r =.51. Evaluations of musician responses for the ‘B’ version of
the etude was in the medium-high range (r =.67) . Once again, estimated reliability scores
were higher for the ‘B’ version than that of the ‘A’ version.
50
Given that these reliability coefficients were gathered from small groups (n = 5)
of conductor evaluations, the r = .51 – .67 range might be considered acceptable
reliability.
Analysis of Data
As described above, the evaluators created a highlighted representation of the
interpretive aspects of the performances as they studied the corresponding markings on
the ‘A’ or ‘B’ version of the excerpt. A 1-5 Likert Score was assigned for each
performance. The conductor scores (GRI Gesture) and musician response scores (GRI
Response) were matched and listed in an excel file.
When examining Hierarchical Linear Models, the unit of measure must remain
independent so as not to violate assumptions. With this in mind, the mean score of the
panel evaluations was used as the unit of measure for GRI Gesture and GRI Response
scores. Each conductor’s GRI Gesture score reflects the average score of the four-
member panel that evaluated conductor video performances (Table 4). These GRI
Gesture scores were used to separate the conductors into Low (n = 5), Medium (n = 5), or
High (n = 5) categories. Table 8 displays this ranking.
51
Table 8
Conductor Ranking: categorized based on GRI Gesture score.
GRI gesture score

CONDUCTOR MEAN
15 4.375
3 3.5
10 3.375
5 3.375
1 3.25
4 3.125
13 3
12 2.625
8 2.625
7 2.5
11 2.25
6 2.125
2 2
9 1.75
14 1.375
High Conductor Group n=5 (M 3.58, SD .46)
Low Conductor Group n=5 (M 1.98, SD .35)
Five null hypotheses were constructed in order to apply an Analysis of Variance
(ANOVA) for these data:
H1: There is no difference between corresponding GRI Musician mean

Response scores of conductors who score high on GRI Gesture and those
conductors who score low on GRI Gesture.
H2: There is no interaction between Baseline musician mean response

scores and corresponding Post (w/Conductor) musician mean response
scores for conductors with High and Low GRI Gesture scores.
H3: There is no interaction between Baseline musician mean response

scores and corresponding Post (w/Conductor) mean response scores for
each excerpt (A&B).
52
H4: There is no difference between GRI conductor mean gesture scores for
the (A) and (B) versions of the excerpt.
H5: There is no difference between brass, woodwind, and string musician

mean response scores for the (A) or (B) versions of the excerpt.
The null hypotheses were developed in order to examine the question of gesture
influence on musician performance. Each null hypothesis was examined using the
appropriate ANOVA design.
The first question examined whether conductors who received higher GRI
Gesture scores were actually able to solicit higher GRI Response scores from the
musicians than those conductors who received lower GRI Gesture scores. A significant
difference between high and low conductor groups indicates that conducting gestures
alone can elicit consistent musical responses from musicians. The first null hypothesis is
stated as follows:
H1: There is no significant difference between corresponding GRI

Musician mean Response scores of conductors who score high on GRI
Gesture and those conductors who score low on GRI Gesture.
Conductors were categorized into high-mid-low levels based on their GRI Gesture
scores. Table 9 displays the main effect means for musician responses sorted by
conductor level.
53
Table 9
Conductor Level (CONLEVEL) and GRI Response scores
CONLEVEL GRI Response

(Hi-Mid-Low) Mean (SD)
Hi (n=5) 3.10, (.47)
Mid (n=5) 2.85, (.44)
Lo (n=5) 2.47, (.36)
Sample mean of all responses & standard deviations (SD) sorted by conductor level
Two versions of the excerpt were utilized. Table 10 displays the means of
musician nests that performed for each conductor group.
Table 10
Conductor Level and GRI Response scores for each version (A&B).
CONLEVEL GRI Resp 'A' GRI Resp 'B'

(Hi-Mid-Lo) Mean (SD) Mean (SD)
Hi (n=5) 2.86, (.54) 3.34, (.41)
Mid (n=5) 2.96, (.46) 2.74, (.42)
Lo (n=5) 2.26, (.40) 2.68, (.33)
Sample mean (M) & standard deviation (SD)
The ANOVA procedure that examined conductor level (high, middle and low)
and excerpt is illustrated in Table 11. Results indicate that conductors of the High
classification (M 3.10) were able to elicit significantly higher GRI Response scores from
musicians than conductors of the Low classification (Mean Difference = 2.47). There is an
interaction between Conductor level and whether the conductor was attempting to
interpret the ‘A’ or ‘B’ version of the excerpt. Conductors of both the Low and High
54
classifications elicited higher GRI Response scores from musicians when attempting to
interpret the ‘B’ version of the excerpt (Table 10).
Table 11
GRI Musician Response: ANOVA Procedure for Conductor Level and Excerpt
Source DF F Value Pr > F

CONLEVEL 2 26.57 < .0001
EXCERPT 1 10.10 .0018
CONLEVEL*EXCERPT 2 10.09 < .0001
Conductors of the mid-ability group (n=5) scored slightly higher when attempting
to interpret the ‘A’ version of the excerpt. A post hoc Tukey test examined pairwise
comparisons between conductor groups and indicates s a significant (p < .05) and large
effect size (.79) between the Low Conductor group and the Middle and High groups on
version ‘A’ of the excerpt. There was also a significant (p < .05) and large effect size (.61)
displayed between the Low Conductor group and the High Conductor group on version
‘B’ of the excerpt. It is important to note that the difference between mean differences of
GRI Response scores for the mid-ability conductors as they conducted the ‘A’ version
(M = 2.96, SD .46) or ‘B’ version (M = 2.74, SD = .42) is small (M = .22) and led to an
interaction between level of conductor and excerpt. (F = 10.09, p < .0001).
Figure 6 presents the interaction effect for GRI Responses of High-Medium-Low
rated conductors as they interpreted the ‘A’ or ‘B’ version of the excerpt.
55
Figure 6. Conductor & Excerpt Interaction for GRI Response Scores
4.0
3.5 Hi
Med
3.0
Low
2.5
2.0
1.5
1.0
0.5
‘A’ Version ‘B’ Version
The GRI Response scores of mid-ability conductors, with respect to excerpt (‘A’
and ‘B’), display an opposite effect. The mid-ability level conductors scored slightly
higher while attempting to interpret the ‘A’ version of the excerpt (2.79 vs. 2.78). In
Figure 7, mid-ability level conductors are classified as a moderating variable (Jaccard,
1998) and are removed from the interaction graph. Figure 7 illustrates the conclusion that
there exist a significant difference between the responses of musicians to the high rated
conductors (High GRI Gesture) and those conductors who were classified in the low
rated group (Low GRI Gesture) (F=26.57,p < .0001). The differences in Means, while
relatively small, are not due to chance. The high conductor group elicited higher musician
response scores, on both versions of the excerpt, than the low conductor group.
56
Figure 7. GRI Response Scores of High and Low Ability Conductors with Excerpt
4.0
3.5 Hi
3.0
Low
2.5
2.0
1.5
1.0
0.5
‘A’ Version ‘B’ Version
An examination of the GRI Response scores of the musicians displayed a
significant difference between the baseline performance (each musician’s personal
interpretation) and the post performance (while watching the conductor). The null
hypothesis was stated as follows:
H2: There is no interaction between Baseline musician mean response scores and
corresponding Post (w/conductor) musician mean response scores for conductors
with High and Low GRI Gesture scores
Table 12 displays the mean scores of baseline and post performances with Conductors
(w/Conductor) as they relate to the Low, Medium, and High category conductors. The
data suggest that the conducting gestures changed the performance of the musicians and
57
that the changed performance corresponded to the conductor’s pre-determined
interpretation (F=188.18, p < .0001).
Table 12
Group means for Conductor Level (CONLEVEL) and Baseline-Post performances

(w/Conductor) of musicians for each excerpt version (A & B).
CONLEVEL Baseline 'A' w/Conductor 'A' Baseline 'B' w/Conductor 'B'

(Hi-Mid-Lo) mean (SD) mean (SD) mean (SD) mean (SD)
Hi (n=5) 1.79, (.48) 2.99, (.99) 1.99, (.50) 3.42, (.78)
Mid (n=5) 2.14, (.33) 2.79, (.81) 1.78, (.54) 2.78, (.64)
Lo (n=5) 1.97, (.42) 2.15, (.42) 2.05, (.65) 2.63, (.72)
Sample means & standard deviation (SD)
Table 13 displays results of the ANOVA procedure applied to conductor level
(CONLEVEL) and change scores between the baseline (BL) and post performances with
conductors (w/Conductor).
Table 13
ANOVA procedure for Baseline - with Conductor (BL-w/Conductor), Conductor Level

(CONLEVEL), and Excerpt

BL-w/Conductor (Time) 1 188.18 < .0001
BL-w/Conductor *CONLEVEL 2 19.02 < .0001
BL-w/Conductor *EXCERPT 1 7.00 .00091
Error (BL-w/Conductor) 146
BL-w/Conductor *CONLEVEL refers to the interaction between baseline (BL)
and post performance scores with conductors (w/Conductor) as musicians performed
while viewing High and Low ability level conductors. ANOVA found that there was a
58
significant difference between the baseline performance (BL) and the post performance
response (w/Conductor) while watching the conductor videos (F 19.02, p < .0001). A
post-hoc Tukey test determined that a significant difference (p < .05) and large effect size
(.64) existed between the Low Conductor group and the Mid-ability conductor group and
also between the Low and High conductor group (.84) for the ‘A’ version of the excerpt.
A significant difference (p < .05) and large effect size (.79) was also evident between the
Low conductor group and the High conductor group for version ‘B’. The interaction
between the Mid-ability and High conductor group was also significant (p < .05) and
displayed a large effect size (.64).
Hypothesis three examined if a significant interaction existed between
conductor’s ability to elicit specific musician responses and each excerpt (A & B).
H3: There is no interaction between Baseline musician mean response scores

and corresponding Post (w/Conductor) mean response scores for each excerpt
(A&B).
The musician’s w/Conductor (Post) response scores increased significantly

depending on whether High, Middle or Low ability conductors were attempting the ‘A’
or ‘B’ version of the excerpt (Table 14).
Table 14
Baseline and POST GRI Response Scores of High and Low Level Conductors
CONDUCTOR LEVEL Baseline GRI Response w/Conductor GRI Response

‘A’ version / ‘B’ version ‘A’ version / ‘B’ version
High 1.79 1.99 2.99 3.42
Middle 2.14 1.78 2.79 2.78
Low 1.97 2.05 2.15 2.64
59
As reflected through the BL-w/Conductor*Excerpt interaction analysis (Table
15), there was a significant difference between excerpts: that is to say, some conductors
elicited higher GRI Response scores while attempting to conduct the ‘B’ version (F = 7.0,
p = .00091).
Table 15
ANOVA procedure for Baseline - with Conductor (BL-w/Conductor), Conductor Level

(CONLEVEL), and Excerpt

BL-w/Conductor (Time) 1 188.18 < .0001
BL-w/Conductor *CONLEVEL 2 19.02 < .0001
BL-w/Conductor *EXCERPT 1 7.00 .00091
Error (BL-w/Conductor) 146
As previously discussed, A post-hoc Tukey test compared the between group
means of the High, Middle and Low ability conductors for significance, finding that the
High and Low group conductors elicited significantly higher GRI Response scores while
conducting the ‘B’ version of the excerpt. The Middle ability conductor group elicited
slightly higher GRI Response scores while conducting the ‘A’ version of the excerpt.
As illustrated in Figure 8 and Figure 9, it is interesting to note that there was a
significant difference between the baseline and post (w/Conductor) performance of
musicians as they performed both versions of the excerpt with High, Medium, and Low
GRI Gesture classified conductors (F = 19.02, p < .0001). Even though the musicians
created only one baseline performance against which all of their subsequent performance-
60
responses were measured, the nested means for the baseline performances differed
between conductors and excerpts. Since the nesting procedure produced varied
combinations of musicians, mean scores for the baseline performances of the nests also
varied. A post-hoc Tukey test revealed a significant (p < .05) yet small effect size (.34)
between the nested Baseline scores, when compared to the ‘A’ version of the excerpt, for
the High and Middle conductor groups (Figure 8). Baseline scores, when compared to
the ‘B’ version of the excerpt, displayed a significant (p < .05) yet small effect size (.26)
between the nested Baseline scores for the Middle and Low conductor groups (Figure 9).
Figure 8. Excerpt ‘A’: Musician Baseline and Post Performance-Responses
BL POST
61
For excerpt A, the largest increase between baseline and post performances
occurred with musician nest that had the lowest baseline score and performed for the high
rated conductors.
Figure 9. Excerpt ‘B’: Musician Baseline and Post Performance-Responses
BL POST
If a significant difference was found between GRI Gesture scores of the two
versions of the excerpt (A & B), it might be concluded that one version was much easier
to conduct or interpret. For the purposes of this study, it was important that both versions
of the etude were similar in terms of interpretive demand, yet required very different
interpretive approaches. In order to determine if there was a difference in difficulty
between versions of the excerpt, the following null hypothesis was examined:
62
H4: There is no difference between GRI conductor mean gesture scores of

the (A) and (B) versions of the excerpt.
As displayed in Table 16, ANOVA concluded that there was no significant difference
between versions ‘A’ and ‘B’ of the etude (EXCERPT) in terms of GRI Conductor
Gesture scores (CONLEVEL).
Table 16
GRI Gesture: ANOVA procedure for Conductor Level and Excerpt (A & B)

CONLEVEL 2 5.51 .0080
EXCERPT 1 2.40 .1238
Error 146
There was no significant difference between the GRI Gesture scores assigned to
each conductor by the panel of conductor evaluators and conductor’s attempts to conduct
the ‘A’ or ‘B’ versions of the excerpt. There was, as already discussed, significant
differences between GRI Response scores of musicians as they performed with
conductors (F = 5.51, p = .008). As discussed earlier, the High and Low level conductor’s
attempts to interpret the ‘B’ version elicited a higher GRI Response from the musicians
(Figure 7).
Table 17 displays the ranked order for all of the conductors based on the product
of the GRI Gesture and GRI Response scores (GRI TOTAL)
63
Table 17
GRI TOTAL: Rank of Conductors High to Low
GRI Gesture GRI Response GRI

CONDUCTOR Gesture score Response score TOTAL
15 4.37 3.43 15.02
10 3.37 3.4 11.47
5 3.37 3.2 10.8
1 3.25 3.15 10.24
3 3.5 2.86 10.03
4 3.12 2.97 9.27
13 3 2.7 8.1
12 2.62 2.97 7.78
8 2.62 2.77 7.26
7 2.5 2.6 6.5
6 2.12 2.66 5.65
11 2.25 2.43 5.47
2 2 2.53 5.07
9 1.75 2.23 3.91
14 1.37 2.23 3.07
An examination of differences between the responses of brass, woodwind, and
string instrumentalists was investigated, either as a confounding variable or as fodder for
future research. Using the following null hypothesis as a directive, no significant
difference between the responses of the three instrument types was found.
H5: There is no difference between brass, woodwind, and string musician

response scores for the (A) or (B) versions of the etude.
Table 18
Means for Instrument type and response per excerpt.
INSTR Excerpt 'A' Excerpt 'B'

mean (SD) mean (SD)
brass 2.73, (1.0) 3.11, (.87)
woodwind 2.54, (.88) 2.86, (.83)
string 2.70, (.75) 2.79, (.41)
Sample mean & standard deviation (SD)
64
Table 18 displays the means and standard deviations for musician responses
sorted by instrument. Results of the ANOVA procedure (Table 19) indicate that there was
no interaction between instrument Type (INSTR) and excerpt A or B (EXCERPT). As
discussed earlier, Low and High group conductors elicited a higher response from
musicians while attempting the ‘B’ version of the excerpt (F = 4.71, p = .0318).
Table 19
ANOVA procedure for Instrument Type

INSTR 2 1.16 .3169
EXCERPT 1 4.71 .0318
INSTR*EXCERPT 2 .32 .7209
Other Observations
Substantial quantities of visual data have been generated through this study.
Video recordings of conductor performances for both versions of the excerpt are quite
revealing, especially comparisons of the High group (top 5) to the Low group (bottom 5).
Highlighted evaluations of the etude displayed the interpretive markings that evaluators
heard musicians bring forth in their performances. There were certain trends that can be
discussed.
65
The vertical and horizontal planes utilized by the High GRI rated conductors
seemed to correspond with rhythmic and lyrical passages. The patterns of the High group
tended to become more vertical while conducting rhythmic passages and more horizontal
while conducting lyrical passages. The lateral plane of the High group conductors tended
to shift as the music increased and decreased in volume and tempi. When there was an
increase in volume or tempi, the lateral conducting plane seemed to move forward. When
there was a decrease in volume or tempi, the plane appeared pulled back.
The High group conductors tended to shift in and out of traditional beat patterns,
conducting more of a general phrase in space and time. This economy of gesture was
evident in the right hand beat pattern and in left hand cues. The Low group conductors
tended to maintain the basic beat patterns throughout the performance. Some
consistencies seemed evident and should be analyzed through future research designed to
appropriately deal with the qualitative data generated through this study.
The highlighted evaluations of the musician responses revealed that certain
articulation patterns, and dynamic tendencies might be intuitive. These intuitive
tendencies were evident in the baseline performances of the musicians, none of whom
collaborated in any way. A research design developed for the analysis of the intuitive
interpretive tendencies of musicians might reveal strong consistencies.
66
Summary
It is important to remember that the musicians had no expressive-interpretive
markings on their excerpt copy. Even so, the results indicate that when musicians
performed the excerpt while viewing high rated conductors, they were able to realize
substantially more of the expressive-interpretive markings than when they performed
while viewing lower rated conductors.
67
CHAPTER V
DISCUSSION
The purpose of this study was to examine the effect of non-verbal conducting
gestures on musicians’ stylistic response and determine if conducting gesture alone can
elicit consistent and accurate musical responses from musicians. It was determined that
some experienced instrumental conductors successfully utilize non-verbal conducting
gestures to communicate specific musical interpretations. It appeared that undergraduate
instrumental music majors (musicians) responded in favorable ways to the musical
interpretation of conductors who had command of a variety of conducting gestures. It
also appeared that those conductors were able to illicit specific musical responses from
the musicians. The results illustrated the existence of a type of social-cultural contract
among musicians: a contract that facilitates non-verbal communication expressed through
gestural conducting.
As discussed in Chapter One, effective conducting requires extensive knowledge
of the musical score combined with an ability to make musical decisions. According to
the results of this study, some of the experienced conductor participants lacked the
gestural technique and vocabulary necessary to convey musical decisions without verbal
instruction. There were however, conducting participants who possessed the technical
ability required to elicit a high degree of targeted response from musicians without the
use of verbal or facial cues.

68
The study of conducting has focused primarily on effective score study
techniques, verbal communication, basic conducting patterns, and facial cues. Few
researchers have examined the effects of gestural technique on performance (Sousa,
1988; Cofer, 1998; Graves, 1984; House, 1998; Sidoti, 1991). As discussed in Chapter
Two, one study developed an instrument that assesses the effect of conducting gestures
on the interpretive response of musicians (Karpicke, 1987). Using the Gestural Response
Instrument developed by Karpicke, this study has confirmed that there exists an unspoken
vocabulary that fine conductors possess. This unspoken vocabulary, when effectively
applied, may have precise influence on the interpretive response of musicians.
As described in the previous chapter, this study involved two groups of
performing participants: conductors and musicians. Conductors were randomly selected
from a pool of fifty experienced conductors, all residing in Central Florida. The
experiences of these instrumental conductors ranged from successful high school
conducting, college-university conducting, to professional band/orchestra conducting.
Quantitative data collected from the assessment of the conductor and musician groups
were analyzed.
Fifteen conductors (N=15) agreed to participate in this study. Two distinct
interpretations of a melodic excerpt were created. Each conductor was asked to conduct
both interpretive versions of the excerpt. These conducting performances were
videotaped and coded with conductor number and excerpt version (‘A’ or ‘B’).
69
Digital technology was used to blur the face of each conductor, thereby eliminating facial
expression and identification.
A panel of four experts evaluated the conductor video performances. These
performances were assessed in two ways. First, the evaluators used a marker pen to
highlight those interpretive items that the conductor conveyed adequately. This process
created a visual representation of the gestural effectiveness of each conductor on each
version of the excerpt. Secondly, the evaluators assigned a 1-5 Likert Score for each
conducting performance. This score was termed as the GRI Gesture score (GRI: Gesture
Response Instrument).
College Musicians from among four college/university music departments
participated in this study. The ability of the musicians varied and represented brass,
woodwind, and string instruments (N=25: 10 brass / 10 woodwind / 5 string).
The musicians were asked to perform from an unmarked edition of the Schubert
melody, Der Jungling Am Bache. Each musician first recorded a performance that
represented a personal interpretation of the excerpt. That initial recording was labeled as
the baseline performance. While viewing selected conductor videos, the musicians were
asked to perform the etude, as they believed the conductor interpreted it. Each
performance was recorded and coded with a musician number-code and the
corresponding conductor-excerpt code. This process created a baseline and post-with
conductor set of performances that corresponded with specific conductor video
performances.
70
Two panels of three evaluators listened to the musician responses (performances
with conductors). The musicians generated a total of 175 recordings of the excerpt. One
panel (Panel A) reviewed 84 recordings; the other panel (Panel B) reviewed 91
recordings. The evaluators used a marker pen to highlight those interpretive items that
they felt the musician conveyed adequately. This process created a visual representation
for the response of each musician, on each version of the etude. The evaluators then
assigned a 1-5 Likert Score for each of these musician responses. This score was termed
as the GRI Response score. Data were subjected to an Analysis of Variance (ANOVA)
technique to determine whether or not there were differences between baseline
performances and the musician responses generated while viewing conductors.
Educational Impact
Based on substantial accomplishments within their performance and conducting
realms, all of the conductor participants involved in this study were experienced
musicians-conductors. However, some were not able to consistently convey precise
musical messages. The conductor participants were given predetermined interpretations
of the melodic excerpt and were asked to communicate those predetermined musical
decisions. The high rated conductors, who possessed a highly developed gestural
vocabulary as it functions within an apparent conductor-musician contract, were able to
convey those predetermined articulations, dynamics, and phrase markings. Even with the
removal of verbal and facial cues, it appears that these conductors had developed the
71
ability to communicate musical decisions through conducting-gestural technique alone.
The existence of a social-cultural contract among musicians and conductors,
involving non-verbal communication may be supported through the results of this study.
Conductors who continually examine, develop, and exploit this conductor-musician
contract may become equipped to manage rehearsal time wisely and gain a higher and
efficient degree of interpretive response from the musicians under his/her baton.
Through conductor and music teacher training curricula, methods designed to
enhance the conductor-musician contract should be developed. With results of this study
in mind, the Elizabeth Green (1961) method of psychological conducting may be a most
appropriate component for conductor training and curricula. The Gestural Response
Instrument (Karpicke, 1987) utilized in this study was a reasonable measure of conductor
effectiveness and musician response. Use of the GRI as a suitable instructional tool in
conductor training is encouraged.
Trends that appeared in the video and evaluative process of this study may
suggest that the isolation of specific “emblems” (Sousa, 1988), the development of a
wide-ranging gestural vocabulary, and proper use of the horizontal, vertical, and lateral
planes appear key in the effective use of the conductor-musician contract.
For the music educator/conductor, the exploitation of conducting technique paired
with strong non-verbal communication may enhance a teacher’s ability to manage music
rehearsal behaviors and develop the conductor-musician contract in young
72
musicians. The research of Cofer (1998) and Sousa (1988) confirm that instruction in the
recognition of specific conducting gestures may enhance the interpretive performance of
young musicians. The conductor-musician contract may be the milieu from which
musical interpretation is communicated non-verbally. Methods that enhance
the development of this conductor-musician contract should be incorporated as a
dimension of ensemble training.
Future Research
As discussed in Chapter Four, the visual data generated though this study is ripe
for further research. Examination of the highlighted evaluations for the musician baseline
performances and responses displayed interpretive trends. It is possible that musicians
have developed intuitive interpretive approaches consistent with certain types of passages
and motives. If this is the case, one can surmise that the interpretive intuition of
musicians is also an aspect of the conductor-musician contract. A study designed to
examine the intuitive approaches of musicians would certainly add to research in music
perception, interpretation, and conducting.
This particular study categorized conductors as High, Medium, and Low gestural
communicators. Consistent differences seemed to exist in terms of pattern, use of plane,
gravity, and “emblems.” A study designed to examine the gestural tendencies of high
scoring conductors could aid in the development of effective methods for training
conductors in the use of gesture and exploitation of the conductor-musician contract.
73
This study focused on non-verbal communication through the baton and hands of
conductors. A study, which focuses on facial cues as non-verbal communication, could
utilize a similar research design and the Gestural Response Instrument (GRI). A
comparison of GRI scores generated through this study and the GRI scores generated
through a facial cues study would aid in the development of research based conducting
curricula and further exploration of the conductor-musician contract.
Although the results of this study are important, caution in the interpretation of
these outcomes is recommended. Researchers who might replicate this design are advised
to prepare the evaluators of conductor videos and musician responses through a more
rigorous procedure. Evaluators involved in this study were given written and verbal
instructions on the highlighting and Likert Scale scoring process. The implementation of
a more specific rubric that corroborates the number of highlighted responses with a Likert
Score rating is recommended (Ex: three highlights = Likert rating of ‘1’; eight highlights
= Likert rating of ‘3’; etc.). Although the reliability quotients of this study are acceptable
given the number of subjects, inter-rater reliability coefficients may rise if evaluators are
exposed to actual practice sessions.
The melodic excerpt used in this study was marked with two different
interpretations: however, the nature of both interpretations of the melody (Der Jungling
Am Bache) was lyrical. Future studies might utilize a wider variety of lyrical and
rhythmic interpretations. The implementation of this research design with individual
musicians, who were instructed to pay careful attention to the conductor’s gestures, may
74
exaggerate the ecological validity of the results. Individual musician performances were
evaluated (rather than ensemble performances) so that the influence of other musicians on
individual interpretation could be controlled. In order to increase ecological validity,
future designs might utilize pairings or small ensembles with one-on-a-part.
No significant difference was found between the responses of brass, woodwind,
and string instrumentalists. Even so, one should be reminded that the sample size of
respective instrument types were not large. A study designed to examine the question of
interpretive response and instrumentalist type might yield different results.
The suggestions outlined in this chapter are not inclusive of all possible related
areas of research that could add to the development of more effective instrumental
conducting methods.
Synthesis of Conducting Domains
The three domains that contribute to effective conducting must function in a
unified way: 1) knowledge and conceptualization of the score, 2) pedagogical ability to
coach toward a common interpretation, and 3) communication of interpretation through
verbal and non-verbal means (Wilcox, 2003). Conductors who excel in all three of these
domains have tremendous impact on the musical effect of ensembles under their baton.
This study has focused on one variable of the communication domain; non-verbal
conducting gestures. It appears that, through gesture alone, certain experienced
conductors possess an extraordinary ability to direct the expressive-interpretive
75
performance of musicians. While all of the conductors involved in this study were able to
perform basic conducting patterns and communicate simple changes in tempi and
dynamics, there were those who communicated far more. The continued study of
effective conducting techniques and gestures, and the development of methods that
exploit the apparent conductor-musician contract, will develop fine conductors who
effectively communicate broad and appropriate musical decisions.
76
References
Berlioz, H. (1970 republished). The conductor: the theory of his art. London: William
Reeves.
Bulwer, J. (1974 republished). Chirologia: or, the natural language of the hand, and
chironomia. Carbondale: Southern Illinois University Press.
Burnsed, V. (2001). Differences in preference for subtle dynamic nuance between
conductors, middle school music students, and elementary school students.
Journal of Research in Music Education 49(2), 49-56.
Campbell, W. & Heller, J. (1979). Convergence procedures for investigating music
listening tasks. Bulletin of the Council for Research in Music Education, 59.
Cofer, R. S. (Fall 1998). Effects of conducting-gesture instruction on seventh-grade band
students’ performance response to conducting emblems. Journal of Research in
Music Education 46(3), 360-73.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ:
Erlbaum Associates.
Dunham, P., Lemke, M. & Moran, P. (1991). Effect of equal and random amounts of
varied practice on transfer task performance. Perceptual and Motor Skills 73,
673-674.
Elliott, D. (1995). Music matters. New York: Oxford University Press.
Fiske, H. (1990). Music and mind. Lewiston, NY: Mellon Press.
77
Fredrickson, W.E. (1992). An exploratory study of band musicians’ eye contact and
performance as influenced by loss of visual and/or aural stimuli. Ph.D. diss.,
Florida State University.
Fredrickson, W. E. (1997). Elementary, middle, and high school student perceptions of
tension in music. Journal of Research in Music Education 45(4), 626-635.
Gallops, W. (1999). Taking note of notes. Teaching Music 7(1), 32-35.
Graves, D. (1984). The choral conductor’s communication of musical interpretation.
Ph.D. diss. University of Connecticut.
Grechesky, R. (1985). An analysis of nonverbal and verbal conducting behaviors and
their relationship to expressive musical performance. Ph.D. diss., University of
Wisconsin. .
Green, E. (1961). The modern conductor. New Jersey: Prentice Hall.
Heller, J. & Campbell, W. (1981). A theoretical model of music perception and talent.
Bulletin of the Council of Research in Music Education. 21-22.
Holahan, J. & Saunders, T.C. (1997). Children’s discrimination of tonal patterns: pattern
contour, response time, and item difficulty level. Bulletin of the Council for
Research in Music Education 132, 85-101.
House, R. (1998). Effects of expressive and non-expressive conducting on the
performance and attitudes of advanced instrumentalists. D.M.A. diss., Arizona
State University.
78
Jaccard, J. (1998). Interaction effects in factorial analysis of variance. Thousand Oaks,
CA: Sage.
Karpicke, H. A. (1987). Development of an instrument to assess conducting gesture and
validation of its use in orchestral performance. Ed.D. diss., University of Houston.
Lewis, D. (1999). Conducting musical shape. Ph.D. diss., Oklahoma State University.
Kromrey, J. & Dickenson, W. (1996). Detecting unit of analysis problems in nested
designs: statistical power and rates of the F test for groups-within-treatments
effects. Educational and Psychological Measurement 56, 215-231.
Madsen, C. & Fredrickson, W.E. (1993). The experience of musical tension: a replication
of Nielsen’s research using the continuous response digital interface. Journal of
Music Therapy 30, 46-63.
Madsen, C. & Madsen, C. (1978). Experimental research in music. Raleigh, NC:
Contemporary Publishing Co..
Madsen, C. & Moore, R. S. (1978). Experimental research in music: workbook in
design and statistical tests. Raleigh, NC: Contemporary Publishing Co..
Mayne, R. (1992). Musical conducting gestures and their interpretation. Ph.D. diss.,
Ohio State University.
Meyer, L. B. (1991). A pride of prejudices; or, delight in diversity. Music Theory
Spectrum 13(2), 241-51.
79
Meyer, L. B. (1976). Grammatical simplicity and relational richness. Critical Inquiry
2(2): 693-761.
Meyer, L.B. (2000). The spheres of music: a gathering of essays. Chicago: University of
Chicago Press.
Miller, S. (1988). The effect of labian movement theory on the ability of student
conductors to communicate musical interpretation. Ph.D. diss., University of
Wisconsin.
Moerbeek, M.,van Breukelen, J & Berger, M. (2000). Design issues for experiments in
multi-level populations. Journal of Educational and Behavioral Statistics 25(3),
271-284.
Rudolf, M. (1950). The grammar of conducting. NewYork: Schirmer.
Schuller, G. (1997).The complete conductor. Oxford: Oxford University Press.
Skadsem, J. A. Effect of conductor verbalization, dynamic markings, conductor gesture,
and choir dynamic level on singer’s dynamic response. Journal of Research in
Music Education 45 n4, (Winter 1997): 508-519.
Sidoti, V.J. (1990). The effects of expressive and non-expressive conducting on the
performance accuracy of selected expression markings by individual high school
instrumentalists. Ph.D. diss., Ohio State University. Dissertation Abstracts
International 51 (3270A).
80
Sousa, G. (1988). Musical conducting emblems: An investigation of the use of specific
conducting gestures and their interpretation by instrumental performers., Ph.D.
diss., Ohio State University. Dissertation Abstracts International 49 (2143A).
Strouse, L. (1987). From analysis to gesture: A comprehensive approach to the score.
Ph.D. diss., Ball State University.
Tighe, T. & Dowling, J. (1993). Procedural and declarative knowledge in music
cognition and education. Psychology and music: the understanding of melody
and rhythm. Hillsdale, N.J.: Erlbaum Press.
VanWeelden, Kimberly (2002). Relationships between perceptions of conducting
effectiveness and ensemble performance. Journal of Research in Music Education
50(2), 165-76.
Wilcox, D. (2003). Unpublished paper presented at the Midwest Band and Orchestra
Conducting Clinic. Chicago, IL.
Zucker, D. (1990). An analysis of variance pitfall: the fixed effects analysis in a nested
design. Educational and Psychological Measurement 50 (Winter), 731-735
81
APPENDICE
i
APPENDIX A: Demographics of Conducting Participants
Conductor education experience gender instrument(s) level exper.
I BME Beg/Adv Male trp HS
II BA Beg/Adv Male tromb/tba C

Sem/Pvt
III BME/MM Beg/Adv/Sem Male tba/pno MS/HS/C/CM/Pro

DMA
IV BME/MM Beg/Sem/Pvt Male cl MS/HS/C
V BME/MME Beg/Adv Female fl MS/HS
VI MME/MA Beg/Adv Male cl/sx/fl MS/HS/C/CM
VII BME/MA Beg/Adv/Sem/Pvt Female tromb HS/C/CM
VIII BME Beg/Adv Male vla/vln MS/HS
IX BME Beg/Adv Male tromb MS/HS
X BME Beg/Adv/Sem/Pvt Female hn MS/HS
XI BA/MM Beg/Sem/Pvt Male trp MS/HS/C/CM
XII BA/MM Beg/Adv/Sem/Pvt Male pno/voc HS/C/CM/Pro
XIII BME/MM Beg/Adv/Sem/Pvt Male tromb MS/HS/C/CM

Ph.D.
XIV BA Music Beg/Adv/Sem/Pvt Male fl/hn HS
BA Music
XV /MME/post Beg/Adv/Sem/Pvt Male clar/pno/voc MS/HS/C/CM/Pro
MME/post
Education: BA/BM/BME=Bachelors Degree in Music or Music Education
MA/MME/MM= Masters Degree in Music or Music Education
Experience: Beg= beginning conducting course / Adv= advanced conducting course / Sem= seminar in
conducting / Pvt=private instruction in conducting
Level Experience: MS=middle school band-orchestra / HS=high school band-orchestra /
C= college band-orchestra / CM=community band-orchestra / Pro= professional band-orchestra
i
APPENDIX B: Musician Nesting With Conductor Videos
Conductor Excerpt Musicians nested to each conductor video

I A B1 B3 W1 W3 S1
B B5 B7 W5 W7 S5
II A B2 B9 W2 W9 S2
B B4 B6 W4 W6 S4
III A B8 B10 W8 W10 S3
B B1 B3 W1 W3 S1
IV A B5 B7 W5 W7 S5
B B2 B9 W2 W9 S2
V A B4 B6 W4 W6 S4
B B8 B10 W8 W10 S3
VI A B1 B3 W1 W3 S1
B B5 B7 W5 W7 S5
VII A B2 B6 W2 W6 S2
B B4 B6 W4 W6 S4
VIII A B8 B10 W8 W10 S3
B B1 B3 W1 W3 S1
IX A B5 B7 W5 W7 S5
B B2 B9 W2 W9 S2
X A B4 B6 W4 W6 S4
B B8 B10 W8 W10 S3
XI A B1 B3 W1 W3 S1
B B5 B7 W5 W7 S5
XII A B2 B9 W2 W9 S2
B B4 B6 W4 W6 S4
XIII A B8 B10 W8 W10 S3
B B1 B3 W1 W3 S1
XIV A B5 B7 W5 W7 S5
B B2 B9 W2 W9 S2
XV A B4 B6 W4 W6 S4
B B8 B10 W8 W10 S3
ii
APPENDIX C: Unmarked Musical Excerpt used in the study.
iii
APPENDIX D: Marked Versions ‘A’ and ‘B’ of the excerpt used in the study.
iv
APPENDIX E: Sample of Marked and Evaluated Excerpts ‘A’ & ‘B’ and Description
musician code evaluator

response rating conductor number
and excerpt
highlight of perceived response
v
APPENDIX F: Musician Response Scores for Each Conductor Nest
Version ‘A’ Version ‘B’

GRI
Musician Response
CONDUCTOR MUSICIAN IA IIA IIIA IB IIB IIIB Response mean Score
1 1 2 4 3 5 5 4 3.428571429
1 2 2 3 2 3 3 4 2.833333333
1 3 2 4 3 3 5 3 3.333333333
1 4 2 3 3 3 3 3 2.833333333
1 5 3 4 2 3 4 4 3.333333333
RespTOTAL
Cond 1 3.152
2 1 2 3 2 1 2 2 2
2 2 2 3 3 4 2 3 2.833333333
2 3 3 5 3 1 2 2 2.666666667
2 4 2 2 2 3 4 4 2.833333333
2 5 1 2 2 3 2 4 2.333333333
RespTOTAL
Cond 2 2.533
3 1 3 5 5 1 4 3 3.5
3 2 3 3 3 2 4 2 2.833333333
3 3 1 2 2 1 3 3 2
3 4 3 3 3 3 3 3 3
3 5 2 4 2 3 4 3 3
RespTOTAL
Cond 3 2.866
4 1 2 4 2 3 4 2 2.833333333
4 2 1 3 2 3 3 4 2.666666667
4 3 1 4 2 5 4 3 3.166666667
4 4 2 3 2 4 3 4 3
4 5 3 4 4 1 4 3 3.166666667
RespTOTAL
Cond 4 2.966
5 1 1 1 1 4 4 5 2.666666667
5 2 5 4 4 4 4 4 4.166666667
5 3 1 3 2 2 2 3 2.166666667
5 4 4 4 3 3 3 4 3.5
5 5 3 4 4 3 4 3 3.5
RespTOTAL
Cond 5 3.2
6 1 2 4 3 2 3 2 2.428571429
6 2 2 4 2 2 3 3 2.571428571
6 3 2 3 3 2 3 2 2.571428571
6 4 1 1 2 4 4 3 2.714285714
6 5 2 4 2 2 3 3 3
RespTOTAL
Cond 6 2.657
vi
APPENDIX F (continued)
7 1 3 5 3 1 2 2 2.666666667
7 2 2 3 2 2 3 4 2.666666667
7 3 2 3 2 2 3 2 2.333333333
7 4 1 3 3 1 3 3 2.333333333
7 5 2 3 2 4 3 4 3
RespTOTAL
Cond 7 2.6
8 1 4 4 4 2 4 3 3.5
8 2 1 4 3 2 4 3 2.833333333
8 3 3 5 3 2 3 4 3.333333333
8 4 1 2 2 2 2 2 1.833333333
8 5 2 3 2 2 3 2 2.333333333
RespTOTAL
Cond 8 2.766
9 1 1 1 1 4 5 3 2.5
9 2 1 2 2 3 3 3 2.333333333
9 3 2 1 1 5 4 2 2.5
9 4 1 2 2 1 1 1 1.333333333
9 5 3 2 3 2 3 2 2.5
RespTOTAL
Cond 9 2.233
10 1 1 2 2 5 4 5 3.166666667
10 2 4 4 4 3 5 4 4
10 3 1 4 3 1 2 3 2.333333333
10 4 4 5 4 5 3 5 4.333333333
10 5 4 3 4 2 4 2 3.166666667
RespTOTAL
Cond 10 3.4
11 1 3 3 2 2 3 2 2.5
11 2 2 3 1 2 3 2 2.166666667
11 3 1 3 3 4 3 2 2.666666667
11 4 1 3 2 3 3 2 2.333333333
11 5 2 4 2 1 3 3 2.5
RespTOTAL
Cond 11 2.433
12 1 5 5 3 2 3 1 3.166666667
12 2 2 2 2 4 4 5 3.166666667
12 3 4 4 3 1 3 3 3
12 4 4 2 3 2 4 3 3
12 5 1 3 2 3 2 4 2.5
RespTOTAL
Cond 12 2.966
13 1 3 5 4 3 3 2 3.333333333
13 2 3 3 4 3 4 2 3.166666667
13 3 2 2 1 1 2 2 1.666666667
13 4 3 4 3 2 3 3 3
13 5 1 3 2 2 4 2 2.333333333
vii
APPENDIX F (continued)
RespTOTAL
Cond 13 2.7
14 1 1 2 2 2 4 2 2.166666667
14 2 1 3 2 2 3 2 2.166666667
14 3 2 2 2 5 3 3 2.833333333
14 4 2 2 2 1 2 3 2
14 5 1 2 2 2 3 2 2
RespTOTAL
Cond 14 2.233
15 1 1 2 2 4 4 5 3
15 2 5 4 4 4 4 4 4.166666667
15 3 1 3 3 3 3 3 2.666666667
15 4 3 4 4 4 5 4 4
15 5 4 4 4 2 3 3 3.333333333
RespTOTAL
Cond 15 3.433
viii
APPENDIX G: Evaluation of POST (w/Conductor) Musician Responses
Panel A: Evaluation of Musician Responses
conductor-
excerpt MUSICIAN evaluator I evaluator II evaluator III
1a 1 2 4 3
11a 1 2 4 3
16a 1 3 3 2
3a 2 2 3 2
12a 2 4 5 3
17a 2 5 5 3
1a 3 2 3 2
11a 3 2 4 2
16a 3 2 3 1
20a 4 1 2 2
10a 4 1 1 1
15a 4 1 2 2
9a 5 2 4 2
14a 5 1 1 1
19a 5 1 2 2
8b 1 2 4 3
13b 1 2 4 3
18b 1 3 3 2
9b 2 3 4 2
14b 2 5 5 3
19b 2 2 4 2
8b 3 2 4 2
13b 3 2 4 3
18b 3 3 4 2
3b 4 1 2 2
12b 4 1 2 2
17b 4 2 3 1
1b 5 5 5 4
11b 5 2 3 2
16b 5 2 3 2
1a 11 2 4 3
11a 11 2 3 3
16a 11 1 3 3
3a 12 3 5 3
12a 12 3 5 3
17a 12 4 4 3
1a 13 2 3 3
11a 13 1 1 2
ix
APPENDIX G (continued)
16a 13 1 3 2
20a 14 1 3 3
10a 14 1 3 2
15a 14 1 4 3
9a 15 1 4 2
14a 15 2 1 1
19a 15 2 2 2
8b 11 1 3 3
13b 11 2 3 4
18b 11 1 2 2
9b 12 5 4 3
14b 12 5 4 2
19b 12 5 3 3
8b 13 3 3 3
13b 13 2 2 2
18b 13 2 3 3
3b 14 1 2 2
12b 14 2 3 2
17b 14 1 3 3
1b 15 3 5 3
11b 15 2 3 2
16b 15 4 3 2
1a 21 3 4 2
11a 21 2 4 2
16a 21 2 4 2
3a 22 1 2 2
12a 22 2 3 2
17a 22 1 3 2
18a 23 1 3 2
8a 23 2 4 2
13a 23 2 3 2
8b 21 3 4 3
13b 21 2 3 2
18b 21 2 4 2
9b 22 1 4 3
14b 22 2 3 2
19b 22 2 3 2
20b 23 2 3 3
10b 23 3 4 3
15b 23 2 4 2
x
Panel B: Evaluation of Musician Responses
cond-excerpt MUSICIAN evaluator IV evaluator V evaluator VI
20a 6 4 3 5
10a 6 4 4 5
15a 6 4 4 4
9a 7 2 3 1
14a 7 2 2 1
19a 7 2 3 1
18a 8 4 5 3
8a 8 5 5 3
13a 8 4 4 4
3a 9 3 3 2
12a 9 2 3 2
17a 9 2 2 2
18a 10 4 3 3
8a 10 3 3 2
13a 10 3 4 1
3b 6 3 2 4
12b 6 4 3 2
17b 6 5 4 4
1b 7 4 3 3
11b 7 3 3 2
16b 7 2 3 2
20b 8 5 4 4
10b 8 5 4 4
15b 8 5 4 5
9b 9 4 3 3
14b 9 3 3 3
19b 9 3 2 1
20b 10 4 4 4
10b 10 4 4 4
15b 10 4 5 3
20a 16 4 4 3
10a 16 3 4 4
15a 16 4 5 4
9a 17 2 3 2
14a 17 2 2 1
19a 17 2 2 2
18a 18 1 2 2
8a 18 2 1 1
13a 18 3 2 1
xi
3a 19 2 2 2
12a 19 3 3 3
17a 19 3 2 4
18a 20 3 4 3
8a 20 3 3 3
13a 20 2 2 1
3b 16 4 4 3
12b 16 3 3 1
17b 16 3 4 2
1b 17 3 3 3
11b 17 3 4 4
16b 17 2 3 3
20b 18 3 3 3
10b 18 3 2 2
15b 18 3 2 1
9b 19 4 3 4
14b 19 1 1 1
19b 19 3 2 1
20b 20 4 5 4
10b 20 4 3 3
15b 20 5 3 5
20a 24 4 4 4
10a 24 4 4 3
15a 24 4 3 4
9a 25 4 4 3
14a 25 3 2 3
19a 25 2 2 1
3b 24 4 2 3
12b 24 4 3 4
17b 24 4 2 3
1b 25 4 4 3
11b 25 3 3 2
16b 25 3 3 1
xii
About the Author
R. Wayne Gallops serves as Director of Instrumental Music Education and Jazz
Studies at Radford University (Virginia). His responsibilities include courses in music
education, conducting, jazz piano, and direction of the RU Jazz Ensembles. Prior to his
appointment at Radford University, Professor Gallops served on the music faculties of
the University of Tampa and Hillsborough Community College. He performs and
conducts throughout the United States and his K-12 public school teaching experience
spans sixteen years. Mr. Gallops holds degrees from Hillsborough Community College,
the University of Tampa, Florida State University, and entered the Ph.D. program at the
University of South Florida in 2001.
61

The Effect of Conducting Gesture On Expressive-Interpretive Perfo

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Effect of Conducting Gesture On Expressive-Interpretive Perfo

Uploaded by

Copyright:

Available Formats

University of South Florida

Graduate Theses and Dissertations Graduate School

The Effect of Conducting Gesture on Expressive-Interpretive

Follow this and additional works at: https://scholarcommons.usf.edu/etd

Part of the American Studies Commons

Scholar Commons Citation

Performance of College Music Majors

Ronald Wayne Gallops

A dissertation submitted in partial fulfillment

Major Professor: Jack Heller, Ph.D.

Keywords: Music, Conducting, Gesture, Interpretation, Perception

 Copyright 2005, R. Wayne Gallops

the Hillsborough County Public Schools, the University of Tampa, Hillsborough

I am blessed with wonderful family members who have on many occasions,

generosity and love of music.

LIST OF TABLES iii

CHAPTER II: Review of Literature 12

CHAPTER III: Research Procedures 29

CHAPTER IV: Data Analysis 38

ABOUT THE AUTHOR End Page

Table 1 Mean performance ratings and standard deviations

Table 2 Mean scores of instructional conditions and dynamic

Table 3 Conducting participant demographics 40

Table 4 Conductor video performance evaluations

Table 5 Evaluations of musician baseline performances 47

Table 6 Musician mean response scores for each conductor 48

Table 7 GRI total scores 49

Table 8 Conductor Ranking: categorized based on GRI gesture score 51

Table 9 Conductor level and GRI response scores 54

Table 11 GRI Musician Response: ANOVA Procedure for conductor level

Table 12 Group means for conductor level and baseline-post

Table 13 ANOVA procedure for baseline-with conductor, conductor level,

Table 16 GRI Gesture: ANOVA procedure for conductor level and

Table 17 GRI Total: rank of conductors high to low 64

Table 18 Means for instrument type and response per excerpt 64

Table 19 ANOVA procedure for instrument type 65

Figure 1. GRI scoring box 27

Figure 2. Nesting procedure 34

Figure 3. Creation of the GRI Scores 36

Figure 4. Melodic excerpt (unmarked) 42

Figure 5. The Evaluation Process: sample of a highlighted copy

Figure 6. Conductor and excerpt interaction for musician GRI response

Figure 8. Excerpt A: musician baseline and post performance-responses 61

Figure 9. Excerpt B: musician baseline and post performance-responses 62

Ronald Wayne Gallops

cues were eliminated, some experienced conductors successfully utilized non-verbal

conducting gestures to communicate specific musical interpretations. It appeared that

command of a variety of conducting gestures. The results illustrated the existence of a

perceptual contract that facilitates the non-verbal communication expressed through

gestural conducting. As demonstrated through this study, some experienced conductors

decisions while others were proficient in this area.

This study examines how conductors utilize non-verbal gesture to communicate

Many factors influence a conductor’s ability to communicate stylistic interpretation and

the musician’s response to the conductor’s communication. Musicians respond to the

musical interpretation of conductors beyond precision of notes and rhythms through a

communication may illustrate a type of social-cultural contract among musicians.

Modern conducting technique, including the use of a baton, is a comparatively

continued increase in complexity and size of symphonic scores created an evolution of

and smoothness of orchestral performance undreamed of even as recently as Richard

Wagner’s time" (Szell in Rudolf, 1950, p.vii).

Three important domains that contribute to effective conducting can be identified:

1) knowledge and conceptualization of the score, 2) pedagogical ability to coach toward a

common interpretation, and 3) communication of interpretation through verbal and non-